# Rampmetrics — Full Content Index Sitemap: https://rampmetrics.com/sitemap.xml This document aggregates the full text of Rampmetrics' public long-form content for LLM ingestion. Each section is a complete article, prefixed with its canonical URL. --- ## Source: https://rampmetrics.com/field-notes/attribution/manifesto The only thing dead about attribution is the name THE WORK IS ALIVE·THE NEED IS PERMANENT·THE CATEGORY IS PROGRESSING·THE WORD IS DEAD·THE WORK IS ALIVE·THE NEED IS PERMANENT·THE CATEGORY IS PROGRESSING·THE WORD IS DEAD· A thesis # The only thing dead about attribution is the name. The loudest voices in marketing measurement are right that something is dead. They're pointing at the wrong corpse. The work isn't dead. The need isn't dead. The category isn't dead. Every marketer still needs to know whether their efforts are paying off and what to do more of. That need is more urgent than ever. What's failing — and what's been failing all along — is the word we gave the discipline that's supposed to serve it. And the failure of the word is causing real harm to the work underneath. SKIP TO THE RENAME → 01 What's actually dead ## Let's bury what should be buried. Before we defend anything, we want to concede everything that deserves conceding. The skeptics aren't crazy. Several specific things in this category really are finished. Naming them honestly is how the rest of the argument earns its standing. THE BIG ONE ### The endless debate about which attribution model is 'right.' First-touch. Last-touch. U-shaped. W-shaped. Linear. Time-decay. Data-driven. Multi-touch. Statistical fingerprinting. They are all valid for different situations. None of them is universally right. The thing that's actually dead isn't any of these models. It's the meeting that derails into arguing about which one to use and walks out without a decision. It's the idea that there's one correct methodology and we just have to find it. What should replace the debate: pick the model that fits the use case, attach a trust rating to whatever number it produces, and move on. Different questions deserve different models. A team trying to understand campaign-level operational performance picks differently than a team trying to credit revenue at the company level. Both are right. The reason this debate persists isn't intellectual confusion — it's that sales is compensated on credit assignment and reads 'let's improve our measurement' as 'let's redistribute the credit.' Marketing usually cares about influence, not sourcing, but the conversation gets dragged toward sourcing because that's where the political weight lives. Until marketing and sales agree, in advance, that measurement isn't credit, the debate has no clean exit. Which is why it never ends. Bury it. ALSO DEAD / Cookie-only data sources A measurement strategy that depended on third-party cookies as its primary signal was always going to break, and it did. Products built on that foundation are finished. The replacement isn't 'no measurement' — it's measurement that doesn't depend on a single fragile signal. ALSO DEAD / First-generation credit-assignment products The early-2010s tools that treated attribution as a credit-allocation game — pick a model, divide the pie, argue about the splits — solved the wrong problem. The work they did is no longer where the value lives, and the products that haven't evolved past it are appropriately on the way out. ALSO DEAD / 'There's one right metric' Sourced pipeline vs. influenced revenue vs. attributable bookings — pick the one that fits the decision you're making. Insisting on a single number that does all jobs equally well is how we ended up with numbers that do no jobs well. Four real deaths. None of them are attribution. They're things that lived inside attribution and are appropriately ending. The discipline they served is more important than ever. 02 What's permanent ## Every marketer needs data to get better. The need underneath all the noise. It doesn't go away when a post goes viral. It doesn't go away when a methodology gets attacked. It's structural to the work of doing marketing in a company that cares whether the work is paying off. Operationalized, it's three questions. QUESTION 01 / THE SCORECARD ### Are we getting better? Is the marketing engine actually improving quarter over quarter, or are we just running it harder? A real answer requires comparable periods, complete data, and the discipline to define 'better' before measuring it. Most teams cannot honestly answer this. The CMO asks it weekly. The CFO asks it quarterly. The answer is usually a confident shrug. QUESTION 02 / THE COMPASS ### What should we do more of? Where should the next dollar go, the next campaign aim, the next quarter's plan focus? The narrow version of the question — 'more of what we've already done' — is the honest one, because it's the only version your data can actually answer. Most teams answer this by intuition, vendor recommendation, or whoever's loudest in the planning meeting. None of those are the data. QUESTION 03 / THE TRUST RATING ### How much should we trust either answer? Every number this software produces should come with a confidence signal. Data completeness, sample size, period comparability, signal-to-noise. The user should know which answers to bet a quarter on and which to treat as directional. Almost no product in this category does this. Every chart looks equally authoritative. That's the actual trust problem. Scorecard. Compass. Trust rating. Three questions, in a loop, forever. Everything in this category is either in service of one of them or it's furniture. 03 The framework ## The Progression. Three stages of operating maturity, plus a parallel discipline. Each stage answers all three questions, at increasing fidelity. The map of where teams are now and where they're going — which is the thing the discourse has been refusing to draw. 1 WALK STAGE 01 2 JOG STAGE 02 3 RUN STAGE 03 L THE LAB PARALLEL Read carefully — The Lab is not Stage 4. It's a different discipline that runs alongside — different inputs, different questions, different cadence. Treating it as the destination is the discourse's biggest distortion. 01 WALK ### Cross-channel campaign visibility. You stop defaulting to first-touch or last-touch alone. You see which campaigns touched the deals that closed, across channels, in one place. The basics — done well. SCORECARD Touched-pipeline counts, period over period. A real baseline for the first time. COMPASS Which channels and campaigns are showing up on closed-won. Coarse, but directional. TRUST RATING Mostly a data-completeness check. Are pipelines clean, are touches landing, is anything missing. HOW TO TELL YOU'RE HERE Your team can answer 'which campaigns touched this closed deal' in under five minutes, in one place, without exporting to a spreadsheet. COMMON FAILURE Treating Stage 1 visibility as the answer instead of the floor. Teams stall here for years and call it 'attribution.' 02 JOG ### Velocity, trend, and segment analysis. You move from 'what happened' to 'what's changing.' Deals moving faster when certain campaigns touch them. Segments accelerating or stalling. The conversation shifts from touch counts to motion. SCORECARD Velocity, conversion rate by stage, and segment-level trend lines that survive period comparison. COMPASS Which segments are accelerating, which are stalling, and which campaigns correlate with the change. TRUST RATING Sample-size and period-comparability checks. 'Is this trend real or is it three deals?' HOW TO TELL YOU'RE HERE Planning meetings reference movement, not just totals. People say 'enterprise is accelerating' before they say 'enterprise has more pipeline.' COMMON FAILURE Confusing correlation with causation. 'Velocity went up when we ran the webinar series' is a story, not a cause. 03 RUN ### Pattern-matched recommendations. You compare current campaign activity against fingerprints of what's worked before. The system tells a story about what to do next, grounded in your actual historical data. Spend decisions reference recommendations, not just opinions. SCORECARD Engine-level health: are campaigns matching the patterns of past winners, or drifting toward past losers. COMPASS Concrete recommendations: 'this campaign matches the fingerprint of the Q2 enterprise winner; double the budget.' TRUST RATING Confidence scores per recommendation, based on similarity to past patterns and signal-to-noise in current data. HOW TO TELL YOU'RE HERE Spend decisions explicitly reference recommendations. The QBR slide shows 'system says X, we did Y, here's the gap.' COMMON FAILURE Treating recommendations as decisions. The system suggests; humans still pick. The team that forgets that gets bitten. L THE LAB ### Causal inference and portfolio modeling. Bayesian marketing mix modeling. Geo holdouts. Incrementality testing at scale. Not a higher rung — a different discipline entirely. It answers a different question, on a different cadence, with different inputs. DIFFERENT INPUTS Aggregated spend by channel and geo, macro indicators, controlled holdouts. Not the touch-level data of Stages 1–3. DIFFERENT EXPERTISE Bayesian statisticians, econometricians, experimental designers. Not the same skill set as the operating team. DIFFERENT CADENCE Quarterly or annual portfolio recalibration. Not the weekly operating loop where Stages 1–3 live. Some teams running well at Stage 3 will commission Lab work occasionally. Some will never need to. Treating the Lab as the destination everyone should reach is the discourse's biggest distortion — and it's what makes the 'attribution is dead, MMM is the future' argument feel revolutionary when it's actually skipping the entire middle. Most working marketers belong at Stage 2 or Stage 3. The discourse is either pretending they're at Stage 1 (and broken) or pretending they should be in the Lab (and aren't). Both are wrong. 04 An honest engagement ## When someone says 'attribution is dead,' what do they mean? Before agreeing or disagreeing, the honest thing is to ask. The answers are revealing — and so is the fact that there are so many of them. #### Do you mean people don't trust the data? Yes — we agree. That's been a real problem for years. Incomplete data, broken pipelines, dashboards that can't be inspected. If that's what's dead, we're nodding along. #### Do you mean teams aren't aligned on how to measure? Also true. Most companies don't have a working definition of what 'good' looks like. Marketing and sales argue about the model rather than the work. That's exhausting, and worth calling out. #### Do you mean the tooling? Which tools? The first-generation credit-assignment platforms? Sure, those are exhausted. The modern multi-touch-with-pattern-recognition stack? Very much alive. Important difference. #### Do you mean a methodology? Which one? Last-touch is finished. Linear attribution was never great. Multi-touch with success-pattern fingerprinting is the most useful it's ever been. These are very different things, and the word 'attribution' doesn't help us tell them apart. #### Do you mean credit assignment as a concept? That's a more interesting argument — but credit assignment is one mechanic inside the discipline, not the discipline itself. The discipline survives without it. The mechanic is what needs to evolve. WHAT WE KEEP FINDING We've asked these clarifying questions a lot. The word 'attribution' means fifteen different things to fifteen different people. That's the problem. Before anyone can productively debate whether attribution is dead, we have to agree on what we're debating. The current vocabulary doesn't let us do that. Which is why renaming this category isn't a vanity project — it's a precondition for having any of these conversations productively. We're going to propose some candidates later, and we want your help picking one. 05 Why this matters ## A bad name is usually harmless. Not this one. Most miscategorized things still work. The acronym CRM is bad in many ways, and customer relationship management happens fine despite the name. Attribution is one of the rare cases where the bad name is actively damaging the work. Three specific harms, observable in the wild. HARM 01 ### It enables the circular argument. The word does double duty as both a mechanic (credit assignment) and a category (the cross-channel revenue measurement discipline). That ambiguity is what lets a respected voice declare the category dead while recommending its actual work. Fix the name and the argument collapses, because the speaker has to pick which thing they mean. Almost every 'attribution is dead' essay survives only on this equivocation. HARM 02 ### It miscalibrates what teams ask for. A marketer who buys 'attribution software' is signaling, through the word itself, that they want credit assignment. The RFP gets written around credit-assignment criteria. The procurement conversation happens in credit-assignment language. The success metrics get set against credit-assignment outputs. Then the software gets used wrong, fails to deliver operational improvement, and the team blames the category. The name shaped the failure before the contract was signed. HARM 03 ### It points at the wrong question. 'Attribution' focuses attention on credit — who deserves it, how to assign it, what the right model is. The right questions, as the framework shows, are whether the engine is getting better and what to do more of. Every time a marketer reaches for 'attribution data,' the word is subtly redirecting them from operational improvement toward credit allocation. The vocabulary is shaping the work, and the work it's shaping isn't the work that matters. This isn't a marketing inconvenience. It's a vocabulary actively producing worse decisions across the industry. Which is why renaming it is not a vanity exercise. 06 What numbers actually matter ## Every metric is either a scorecard or a compass. Sourced pipeline. Influenced revenue. Marketing-attributable bookings. CAC. Pipeline-to-spend ratio. The list grows every quarter, and most of it is either a scorecard metric (telling you whether you're getting better) or a compass metric (telling you what to do next). The harm comes from using one for the other. Sourced pipeline is a fine scorecard at the engine level. It's a terrible compass at the campaign level. Influenced revenue is a useful compass for which segments to lean into. It's an awful scorecard for whether sales is improving. CAC is a portfolio metric, not a campaign metric, and most teams use it the wrong way. The framework forces the question: what stage are you operating from, and is this metric a scorecard or a compass for that stage. The metrics page works through every common one and labels it. READ THE METRICS FRAMEWORK → Our position, in one sentence Pick metrics that match the question you're trying to answer and the stage you're operating from — not the stage you wish you were at, and not the stage the loudest voice in the room operates from. 07 The funeral, and the replacement ## So what do we actually call it? By this point in the page, the case is made: the work is alive, the need is permanent, the framework exists, the discourse is muddled, and the name is causing real harm. Time to replace it. Revenue intelligence beat 'sales analytics' because it described the outcome. This category needs the same move. Below are the candidates we keep coming back to. Vote. Argue. Suggest a better one. Pick as many as fit — vote for every candidate you'd live with. 01 Marketing Performance Analytics Honest. Boring. Describes the work. The performance frame matches what teams actually need: a scorecard for whether the engine is improving and a compass for where to point next. VOTE · 47 02 Marketing Intelligence Borrows the move that worked for "revenue intelligence." Carries the right connotation: pattern recognition, signal extraction, recommendation. Risk: too generic without a qualifier. VOTE · 31 03 Revenue Marketing Intelligence Most precise of the bunch. Names the input (marketing), the outcome (revenue), and the discipline (intelligence). Mouthful. Probably too long for everyday use. VOTE · 22 04 Campaign Intelligence Narrower than the others — focuses on the unit of work most teams actually plan around. Strong inside operational conversations, weaker at the executive layer. VOTE · 14 05 Go-to-Market Analytics Bigger tent — includes sales motion, not just marketing. Useful for orgs where the line between marketing and sales has blurred. Risk: too broad to mean anything specific. VOTE · 9 Lock in your vote Selections above aren't counted until you submit them with an email. We use the email only to dedupe and let our team update the tally — it's never stored on our servers. PICK AT LEAST ONE → SUBMIT A NEW CANDIDATE PROPOSE → Reader-submitted candidates go to our team for review before joining the tally. Live tally 123 VOTES CAST Marketing Performance Analytics 47 Marketing Intelligence 31 Revenue Marketing Intelligence 22 Campaign Intelligence 14 Go-to-Market Analytics 9 Tally is local to your browser for now. Selections above aren't counted until you submit them with an email. 08 Push back ## Write in the margin. Disagree with a stage definition. Argue the Lab really is a fourth rung. Defend last-touch. Tell us we missed a harm. This page improves when you push. LEAVE A COMMENT VisibilityPublic — publish here if our team approves it.Private — just a message to the team. Never published.Public comments are reviewed before publishing. POST → LENA R., DEMAND GEN LEAD The three-questions framing is what I've been trying to articulate for two years. Scorecard, compass, trust rating. Every vendor pitch I sit through pretends to do all three and actually does maybe one and a half. The trust rating piece is the missing one. ANONYMOUS, B2B SAAS I buy the relocation of the Matt argument, but be careful: not all of the 'attribution is dead' crowd is making the same argument. Some really do mean 'measurement is impossible now,' which is a different and worse claim. Worth distinguishing. MARCUS T., AGENCY Where does small-scale incrementality fit? Lift studies on individual channels, not full geo holdouts? Feels like it bridges Stage 3 and the Lab. Probably worth a footnote on the framework. House rules Disagreement welcome. Bad-faith dunks aren't. If you're going to call something dead, say what you'd put in its place. Quote yourself, not influencers. We will edit this page in response to good arguments. Returning to where we started ## The only thing dead about attribution is the name. The work is alive. The need is permanent. The category is progressing. The framework exists. The metrics matter. The trust problem is solvable. And the word — the single five-syllable noun at the center of all of it — is the part that needs to go. — THE RAMPMETRICS TEAM · UPDATED IN PUBLIC More in this series ### Field Notes — a working publication. YOU ARE HERE THE MANIFESTO The only thing dead about attribution is the name. COMPANION Where Are You on the Progression? READ → COMPANION Is Marketing Mix Modeling For You? READ → HISTORY A Brief History of Regression in Marketing. READ → APPENDIX Sources & References. READ → --- ## Source: https://rampmetrics.com/field-notes/attribution/mmm Is Marketing Mix Modeling For You? | Rampmetrics WHAT IS MMM?·IS IT FOR ME?·WHO IS IT ACTUALLY FOR?·AM I MISSING SOMETHING?·DO I NEED A DATA SCIENTIST?·WHAT IS MMM?·IS IT FOR ME?·WHO IS IT ACTUALLY FOR?·AM I MISSING SOMETHING?·DO I NEED A DATA SCIENTIST?· For the curious You've heard marketing mix modeling mentioned as the future. You're wondering whether you should care. We're wondering some of the same things. Let's work through it together. # Is marketing mix modeling for you? An honest exploration, written for B2B marketers who don't have a data science team, don't run experiments at geo-level scale, and have read enough hot takes to wonder if they're missing something. We'll walk through what MMM actually is, how it differs from what most marketers do today, and help you figure out whether it belongs in your work right now. We have a working hypothesis. We're trying to hold it loosely. Before we begin We make software in the marketing performance space. Today, our product does touch-level analysis with statistical pattern matching — which is a different exercise from MMM, as we'll get into. We're actively exploring what role MMM should play in our product and in the broader B2B marketing world, and we're using this document partly to gather feedback and ideas from the people who'd actually use it. We have early hypotheses. We're holding them loosely on purpose. So consider this a thinking-out-loud document, not a verdict. We'll lay out what we know, where we're uncertain, and how you might figure out where you stand. If you have a perspective — whether you've run MMM, considered it, or written it off — we'd like to hear it. 01 The thing itself ## What MMM actually is. Marketing mix modeling is a statistical technique. At its core, it's a regression model — the same kind of math that asks "if I change this, how much does that change?" Specifically, MMM takes your historical marketing spend across channels, your historical revenue or sales, and a bunch of other variables that influence outcomes (seasonality, pricing, the economy, competitor activity), and tries to estimate how much each channel contributed to results. The output looks something like this: "Across the past two years, every dollar you spent on paid social produced roughly $1.40 in attributable revenue, with high confidence. Every dollar on display produced about $0.60, with much lower confidence." Modern Bayesian MMM dresses this up with probability distributions and prior beliefs about how channels behave — saturation curves, decay rates, that kind of thing — but the underlying exercise is the same. The technique was developed for consumer packaged goods, where companies spend across TV, radio, print, and digital, and want to know how to allocate the next quarter's budget across that portfolio. It's been around since the 1960s. The Bayesian variants and the open-source tools (Google's Meridian, Meta's Robyn) are recent, but the core idea is decades old. A useful mental model MMM is what you'd build if you were trying to answer a CFO question: "At a portfolio level, where is our marketing budget paying off, controlling for everything else that's happening?" It's not built to answer "which campaign should I run next week?" or "is this specific account engaging?" Those are different questions, with different tools. 02 The distinction that matters ## Two ways of looking at the same marketing. Most discourse about MMM versus attribution treats them as competing answers to the same question. They're not. They're different exercises looking at different units of analysis, answering different questions. Neither is better. Knowing which one fits your work is the whole game. APPROACH A Touch-based Analysis at the entity level UNIT OF ANALYSIS Individual prospects, accounts, and deals. Every entity carries its own history of touches and interactions. QUESTION IT ANSWERS WELL "Which campaigns touched the deals that closed? What's the pattern of successful accounts? Where is this segment moving?" WHAT IT CAN PROVE Correlation with traceability. You can drill from a closed deal down to the email that prospect opened in February. The data is concrete and inspectable. WHAT IT CAN'T DO Tell you "shift 30% of budget from display to programmatic audio." It doesn't speak portfolio-allocation language. CADENCE Continuous. Refreshed with every new touch, every new deal. Operating-pace data. WORKS BEST FOR B2B, ABM, considered-purchase categories, anywhere the work involves specific accounts and specific campaigns over longer cycles. APPROACH B Aggregate (MMM) Analysis at the population level UNIT OF ANALYSIS Time periods (usually weeks). Individual interactions disappear into channel totals. You're working with sums, not events. QUESTION IT ANSWERS WELL "How does our total spend on each channel correlate with revenue, controlling for everything else that's happening?" WHAT IT CAN PROVE Statistical relationships at the channel level, with uncertainty intervals. Sophisticated correlation with controls — not causation. (Causation requires actual experiments.) WHAT IT CAN'T DO Drill into any specific account or deal. Tell you which campaign in your current quarter is working. Speak to ABM, because aggregation has thrown the account-level data away. CADENCE Quarterly or semi-annual studies. Strategic-pace data. Used to set portfolio direction, not run daily operations. WORKS BEST FOR High-volume consumer marketing, broad-channel portfolios with meaningful TV/radio/OOH spend, organizations doing big strategic reallocation decisions. ABOUT THE "CAUSATION" CLAIM One thing worth being honest about: MMM is sometimes presented as the rigorous, causal approach to marketing measurement, in contrast to "unscientific" touch-based work. That's a bit overstated. MMM is also correlational. It's a regression model. It estimates statistical relationships, controlling for what you put into the model. It's more rigorous than naive last-touch attribution, certainly. But to prove that a marketing input caused a revenue outcome, you need actual experiments — geo holdouts, audience holdouts, incrementality tests. That's a separate methodology, and it sits above MMM on the causal hierarchy, not below it. The honest framing: touch-based analysis is correlation at the entity level. MMM is correlation at the aggregate level, with statistical controls. Incrementality testing is causation, for the specific bet you tested. Three tools, three jobs. None of them is the apex of measurement. They're complementary. MMM isn't a more rigorous version of touch-based analysis. It's a different exercise, looking at a different unit, answering a different question. The choice between them isn't about sophistication. It's about fit. 03 The conditions that make it work ## When MMM is genuinely useful. MMM is a real tool with real strengths. We're not arguing against it. We're trying to be honest about the conditions under which it produces trustworthy, actionable answers — and the conditions under which it doesn't. CONDITION 01 Data volume MMM needs enough observations to estimate channel effects with reasonable confidence. The conventional minimum is 2-3 years of weekly data, which gets you 100-150 observations. For B2B with smaller volumes and noisier signals, you typically need more, not less. A 10,000-leads-per-month B2B company has data, but the signal-to-noise is harder than for a consumer brand with 10,000 daily sales. CONDITION 02 Channel diversity MMM is most valuable when you're spending meaningfully across multiple channels and trying to optimize the mix between them. A B2B company with 70% of spend in paid search and LinkedIn doesn't have much of a portfolio to optimize. MMM applied there produces wide confidence intervals on the small channels and tells you what you already knew about the big ones. CONDITION 03 Sales cycle length MMM works best when the lag between marketing input and revenue outcome is short and consistent. Consumer purchases happen in days or weeks. B2B sales cycles often run 3 to 18 months, and they're variable. The longer and lumpier the cycle, the harder it is for MMM to associate spend with outcomes confidently. This is structurally why MMM came from CPG, not B2B. CONDITION 04 Organizational capacity to act MMM produces portfolio-level recommendations: "shift 25% from display to programmatic audio." Acting on that requires an organization that can pull those levers at that resolution. Most B2B marketing teams can't reallocate quickly across channels at that scale — in-flight campaigns, agency relationships, sales-and-marketing alignment, brand considerations all bind. The recommendation arrives as an abstraction the org can't operationalize. CONDITION 05 Frequency of strategic decisions MMM produces quarterly or semi-annual outputs. If your team is making big portfolio decisions on that cadence, the timing fits. If you're optimizing at campaign-level cadence — adjusting weekly, reading dashboards daily, iterating on creative — MMM isn't speaking your operating language. Different tool for different cadence. CONDITION 06 Resources for the modeling work Historically, MMM has required data scientists or external consultants to handle data prep, model specification, and interpretation. The math itself is software; the judgment around the math has been human. This is the condition AI may genuinely be changing — and the place we're holding our view most loosely. More on that below. 04 Where we're least certain ## Could AI bring MMM down to everyone? This is the part of the conversation where our working hypothesis is genuinely open. We have a view. We could be wrong. For most of MMM's history, the binding constraint hasn't been the math. The math has been productizable for years — Google and Meta have open-sourced their tools, and any technical team can pick them up. The binding constraint has been the judgment around the math: choosing variables, setting priors, transforming data, interpreting outputs, translating recommendations into operational action. That's where the data scientists earn their fees, and that's why MMM has been a Fortune 500 capability rather than a mid-market one. AI plausibly changes this. Modern models are quite good at the kinds of judgment tasks MMM has needed humans for — picking sensible priors from industry benchmarks, flagging when a model is overfitting, translating coefficient outputs into plain-language recommendations. A product that wraps MMM in AI assistance could, in principle, bring it down to teams that couldn't have run it before. They might not even need to know what's happening under the hood. That's the optimistic case, and we don't dismiss it. If MMM becomes a button rather than a project, the calculus genuinely changes — and probably faster than most of us expect. But — and this is the part of our view we hold more firmly — the structural conditions don't go away because the software gets easier. A B2B company with 8,000 leads a month, two main channels, a nine-month sales cycle, and a marketing team of four people won't get meaningfully more value from MMM no matter how seamless the tool gets. The data volume wasn't the binding issue. The fit wasn't there to begin with. Easier MMM helps the teams who already had the structural fit and were blocked by complexity. It doesn't manufacture fit where none existed. Our working hypothesis AI will productize the math layer of MMM over the next few years. More teams will have access to it. We'll likely add it to our own product when our customers want it. But MMM is still a portfolio-question tool, and most B2B marketing isn't a portfolio question. It's an operational one. We could be wrong about the pace. We don't think we're wrong about the fit. 05 Figure out where you stand ## Is MMM for your team? Six questions about how your marketing actually operates. The verdict at the end is our honest read based on your answers — not a sales pitch, just a working assessment. QUESTION 01 How many leads or marketing-influenced opportunities does your team process per month? Under 2,0002,000 – 10,00010,000 – 50,00050,000+ QUESTION 02 How many channels are you spending meaningfully on (more than 10% of total budget)? 1 or 2 channels — most spend concentrated3 to 4 channels — modest diversification5 or more channels including offline/brandNot sure / hard to characterize QUESTION 03 What's your typical sales cycle from first touch to closed deal? Days to weeks (e-commerce, self-serve)1 to 3 months3 to 9 months9 months+ (enterprise B2B) QUESTION 04 When you make big spend decisions, what kind of decisions are they? Campaign-level — what to run next, where to pushSegment-level — which audiences to lean intoPortfolio-level — shifting big % of budget across channelsA mix of all of the above QUESTION 05 Do you have a data analyst or budget for ongoing modeling work? No — small team, no dedicated analystPart-time access to analytics supportYes — dedicated analyst or external partnerOpen to a productized tool that doesn't need one QUESTION 06 What does your team feel like it's missing most from current analytics? Confidence that the data we have is trustworthyA clear read on what's working at the campaign levelPortfolio guidance — "spend more here, less there"Confirmation that big strategic bets caused the lift 06 Where we land ## Our honest read, held loosely. We think MMM is a real, useful tool for the specific class of organizations that have the data volume, the channel mix, the portfolio-level decision cadence, and the operational capacity to act on its outputs. Historically, that's been a Fortune-500-sized club. AI is plausibly opening the door to more teams, and that's good — more tools accessible to more marketers is unambiguously progress. But we don't think MMM is the future for everyone, and we're skeptical of the discourse that pitches it that way. Most B2B marketing teams operate on smaller data, longer cycles, and campaign-level cadences that MMM isn't built for. For those teams, the touch-based, entity-level, operationally-paced work is where the actual value lives — and it's where the trust and clarity gaps in this category most need to be closed. We'd rather help most B2B marketers get really good at the work they're actually doing, with data they can trust, than encourage them to chase a tool that was built for a different shape of company. If MMM becomes genuinely useful for our customers — productized, embedded, useful without requiring a PhD — we'll add it. Until then, we'd rather be honest about what fits where. In a sentence MMM is a real tool for a real class of company. Most B2B marketing teams aren't that class of company. AI may change that. We're watching. Back to the main argument This document is a companion to The Manifesto — our working framework for how teams of every size can get better at marketing measurement without waiting for a future that may not be theirs to inhabit. If MMM is for you, great. If it isn't, the work in front of you is still worth doing well. — THE RAMPMETRICS TEAM · A COMPANION DOCUMENT More in this series Field Notes — a working publication. YOU ARE HERE Is Marketing Mix Modeling For You? An honest exploration with a self-assessment. THE MANIFESTO The only thing dead about attribution is the name. The main argument of the publication. COMPANION Where Are You on the Progression? A self-placement on the five stages most teams operate at. COMPANION A Brief History of Regression in Marketing 140 years of method, 50 years of marketing application. REFERENCES Sources & References The historical record behind the publication's claims. --- ## Source: https://rampmetrics.com/field-notes/attribution/regression A Brief History of Regression in Marketing | Rampmetrics GALTON 1885·PEARSON 1896·FISHER 1922·CPG ECONOMETRICS 1970s·MMA & HUDSON RIVER 1989–1990·DIGITAL ATTRIBUTION 2010s·ROBYN 2020 · MERIDIAN 2025·GALTON 1885·PEARSON 1896·FISHER 1922·CPG ECONOMETRICS 1970s·MMA & HUDSON RIVER 1989–1990·DIGITAL ATTRIBUTION 2010s·ROBYN 2020 · MERIDIAN 2025· Context for the curious Where these statistical methods came from, how they ended up in marketing, and why a 140-year-old idea is suddenly relevant to a B2B CMO. Show your work View all sources → # A brief history of regression in marketing. The statistical method underneath modern marketing measurement is roughly 140 years old. The application of it to marketing is roughly 50. The recent open-source democratization of it is roughly five. Knowing the arc helps explain why some current arguments — including the one declaring attribution dead — sound novel but aren't, and why the methods now reaching B2B marketers have been refined for decades inside an industry most of us don't work in. Brief, accurate, and aimed at CMOs and marketing leaders, not statisticians. 01 The math itself ## A peasant and a peapod. The word "regression" doesn't come from marketing. It comes from Victorian-era genetics — specifically, a observation about peas and human heights that turned into one of the foundational techniques of modern statistics. In the 1880s, the British polymath Francis Galton was studying inheritance — how traits pass from parents to offspring. He noticed something counterintuitive: when tall parents had children, the children were on average shorter than the parents. When short parents had children, the children were taller. The extremes regressed toward the average of the population. He published the observation in 1885 and 1886, and the term he used to describe it — "regression toward mediocrity" — stuck. What Galton didn't fully realize was that the line he'd drawn through his height data — the line that best summarized the relationship between parents and children — was the seed of a much bigger idea. Over the next several decades, Karl Pearson formalized the related concept of correlation (1896), G. Udny Yule extended the math to handle more variables (1907), and Ronald Fisher synthesized the modern regression model in 1922 — building on Carl Friedrich Gauss's earlier method of least squares from the early 1800s. By the mid-1920s, regression analysis was a recognized statistical method: a way to estimate the relationship between a dependent variable (what you're trying to explain) and one or more independent variables (the factors that might explain it), while quantifying how much of the variation is captured by your model versus left to chance. The math hasn't fundamentally changed since. A note worth making Several of the statisticians who developed these methods — Galton, Pearson, and Fisher in particular — were also active proponents of eugenics, and some of their statistical work was originally developed in service of those ideas. The math they produced is sound and is now used responsibly across science and industry. The history is worth knowing honestly. The methods stand on their own; the people who developed them had views we don't share. 02 From economics to soap ## The detour through econometrics. Regression didn't go straight from statistics to marketing. It went through economics first — and the marketing version inherited the shape of the economic version, which is part of why it works the way it does. In the 1930s and 40s, economists Jan Tinbergen and Ragnar Frisch developed econometrics — the application of statistical methods to economic data, particularly the use of regression to model how economic variables relate to each other over time. They shared the first Nobel Prize in Economic Sciences in 1969 specifically for this work. The point isn't the Nobel — it's that by the late 1960s, the idea of using regression on time-series data to estimate causal-looking relationships between aggregate variables was a fully developed, academically respected discipline. Marketing scholars and practitioners noticed. In the 1960s and 1970s, they began applying econometric methods to a specific problem: how much of a CPG brand's sales could be attributed to advertising spend, versus pricing, versus promotions, versus seasonality, versus everything else? Statisticians at the University of Chicago and elsewhere built the first formal marketing mix models in the 1970s. Building one took months. The work was done by econometricians, not marketers. The 1980s saw early commercial adoption — Coca-Cola is often cited as one of the first major brands to use MMM rigorously, and Kraft, Procter & Gamble, AT&T, and Pepsi followed. The pioneers of commercial MMM as a service emerged at the end of that decade: Hudson River Group in 1989 and Marketing Management Analytics (MMA) in 1990. These firms productized what had been a bespoke consulting exercise, and Nielsen and IRI eventually bundled MMM into their standard data contracts. For about thirty years, this was a Fortune 500 capability — accessible only to companies that could afford the data, the consultants, and the year-long modeling cycles. Why CPG, specifically CPG had three things in abundance that made early MMM possible: huge data volumes from store sales tracking (Nielsen, IRI), broad channel diversity (TV, radio, print, in-store), and short purchase cycles where the lag between ad and sale was measurable in days or weeks. B2B has roughly none of those things. Which is why MMM was developed somewhere else, refined for decades in a domain that suits it, and only recently started to be adapted for environments it wasn't designed for. The genetic code is CPG. 03 The digital interlude ## Then the internet happened. For a couple of decades, marketing measurement took a different path — one that didn't need regression at all. That detour shaped a generation of marketers, and it's the reason much of the current discourse exists. When digital advertising emerged in the late 1990s and exploded in the 2000s, it brought a seductive new capability: deterministic tracking. You could see, with apparent certainty, which ad a user clicked, which page they landed on, and whether they bought. The need for statistical modeling seemed to evaporate. Why estimate the relationship between spend and sales when you could just watch the user do it? This was the era of attribution as an industry name — touch-based, click-based, cookie-based credit assignment. First-touch, last-touch, multi-touch, U-shaped, W-shaped. Vendors built products around assigning credit to specific interactions. The aggregate, statistical approach of MMM looked old-fashioned. It was tied to TV and offline, not the dynamic, trackable digital world. For about fifteen years, MMM was something CPG dinosaurs did. Everyone else was tracking clicks. The detour ended for reasons that had nothing to do with statistics: cookie deprecation, Apple's App Tracking Transparency, GDPR and other privacy regulations, walled gardens that stopped exposing user-level data, and the realization that deterministic tracking had always been more deterministic-looking than deterministic-actually. Suddenly, the aggregate statistical methods that had been around since the 1970s started looking relevant again — not as a replacement for digital tracking, but as a companion to it. 04 The recent democratization ## From Fortune 500 to GitHub. In the last five years, regression-based marketing measurement has gone from a Fortune 500 consulting engagement to an open-source library anyone can download. That shift is the reason this conversation is suddenly reaching B2B. In 2020, Meta released Robyn, an open-source MMM library written in R. For the first time, the math that had been guarded behind consulting fees was free to download. Robyn used ridge regression and adstock modeling and was specifically designed to let advertisers measure the contribution of their media spend, including spend on Meta's own platforms — which was Meta's commercial motivation for releasing it, though the tool itself was genuinely useful regardless. In 2022, Google released LightweightMMM, a Bayesian MMM library in Python. In January 2025, Google followed with Meridian, the official successor to LightweightMMM and Google's more advanced, actively-maintained open-source MMM framework. PyMC Labs built PyMC-Marketing, a Python Bayesian MMM library with a strong open-source community. Uber contributed Orbit KTR. (Meta has since discontinued active development of Robyn; Meridian and PyMC-Marketing are now the two most actively maintained options.) What changed isn't the math. The math has been sitting there since the 1920s, and the marketing application since the 1970s. What changed is access. A technique that required a $250,000 consulting engagement in 2010 is now a free download. AI assistance is making the surrounding judgment work — data preparation, model specification, interpretation — increasingly tractable for teams that don't have a data scientist on staff. The democratization is real, and it's relatively new. An important caveat "Open source" does not mean "easy to use." Downloading Meridian or PyMC-Marketing gets you the math. You still need data prep, statistical judgment, and someone who can interpret what the model produces. Forrester and others note that successful open-source MMM programs still require strong in-house data engineering and data science capacity, or a services partner. The democratization is at the math layer, not the operating layer — though that's changing too. 05 The arc, in one view ## 140 years, in a column. 1885–1886Galton coins "regression." Francis Galton observes "regression toward mediocrity" in his studies of inheritance, and invents the first crude form of linear regression. 1896–1907Pearson and Yule formalize the math. Karl Pearson develops correlation theory. G. Udny Yule extends regression to handle multiple variables. The framework starts to look modern. 1922Fisher synthesizes modern regression. Ronald Fisher unifies Gauss's least squares with the Pearson/Yule correlation framework. The math underneath all modern regression analysis is essentially in place. 1930s–1960sEconometrics emerges. Tinbergen and Frisch develop econometrics — the application of regression to economic time-series data. They share the first Nobel Prize in Economic Sciences in 1969. 1960s–1970sMarketing scholars adapt the methods. Researchers begin applying econometric methods to marketing data, primarily in CPG. The University of Chicago is one center of early work. Building a model takes months. 1980sCommercial MMM in major CPG brands. Coca-Cola, Kraft, P&G, AT&T, and Pepsi adopt MMM as part of marketing planning. The method requires a dedicated team of econometricians. 1989–1990MMM becomes a commercial service. Hudson River Group (1989) and Marketing Management Analytics (1990) launch as the first MMM-as-a-service firms. Nielsen and IRI eventually bundle MMM into their data contracts. 2000sThe digital attribution detour. Digital advertising's deterministic tracking promises individual-level attribution. The aggregate statistical methods of MMM look old-fashioned. A generation of marketers learns to think in clicks, not coefficients. Late 2010sThe detour breaks down. Cookie deprecation, ATT, GDPR, walled gardens, and the realization that deterministic tracking was never as deterministic as it looked. MMM and other aggregate methods come back into the conversation. 2020Robyn — the first major open-source MMM. Meta releases Robyn under an open-source license. For the first time, advertisers outside the Fortune 500 can download and run MMM-class statistical models for free. 2022–2024LightweightMMM, PyMC-Marketing, Orbit KTR. Google's LightweightMMM (2022), PyMC Labs's PyMC-Marketing, and Uber's Orbit KTR expand the open-source ecosystem. Bayesian MMM becomes the modern default. January 2025Meridian launches. Google releases Meridian as the official successor to LightweightMMM — Bayesian, geo-level, with integrations for Google Search and YouTube data. Mainstream attention follows. 2025–2026AI accelerates the democratization. AI assistance starts to make the surrounding judgment work — data prep, model specification, interpretation — tractable for teams without dedicated data scientists. The math layer was solved decades ago; the operating layer is solving now. 06 What the history tells us ## So why does this matter to a B2B CMO? A few things worth taking from the arc. First — these methods are old, well-tested, and not faddish. When someone pitches you regression-based marketing analytics in 2026, they're not selling something new. They're selling the application of a hundred-year-old statistical framework that's been refined inside the most rigorous quantitative marketing organizations in the world for fifty years. The novelty is the accessibility, not the math. Second — the methods were developed in CPG, not B2B. The genetic code is short-cycle, high-volume, broad-channel consumer marketing. B2B has different shape — longer cycles, smaller volumes, narrower channel mix. The methods adapt, but the adaptation is genuine work. Treating CPG MMM as a turnkey B2B solution is a category error. Third — the digital attribution era was the detour, not the destination. Cookie-based, deterministic, individual-user tracking was a fifteen-year experiment that's now ending. The methods coming back into the conversation aren't new — they're the older, more durable methods that predate the detour. If you've been in marketing long enough to remember when MMM was something only big CPG brands did, that's where the field is heading again, with better tools. Fourth — the democratization is recent and ongoing. Most of what's now possible for a B2B team to do with these methods wasn't possible five years ago. That's worth knowing, because some of the current discourse — including the "attribution is dead, MMM is the future" argument — is reacting to capabilities that genuinely didn't exist when most of today's marketing leaders learned their craft. The skepticism that comes from "we tried MMM in 2010 and it didn't work for us" is honest, but the world has changed. In one sentence The math is 100+ years old. The marketing application is 50+ years old. The democratization is 5 years old. The B2B fit is still being worked out. Where to next This document is background — the historical context for the rest of the Field Notes publication. If you've read this far and want the operational arguments: the main hub makes the case that the category is misnamed, the Progression companion helps you place your team on the maturity map, and the MMM companion goes deeper on whether marketing mix modeling specifically is right for your situation. More in this series Field Notes — a working publication. YOU ARE HERE A Brief History of Regression in Marketing 140 years of method, 50 years of marketing application, 5 years of open access. THE MANIFESTO The only thing dead about attribution is the name. The main argument of the publication. COMPANION Where Are You on the Progression? A self-placement on the five stages most teams operate at. COMPANION Is Marketing Mix Modeling For You? An honest exploration with a self-assessment. REFERENCES Sources & References The historical record behind the publication's claims. --- ## Source: https://rampmetrics.com/field-notes/attribution/progression Where Are You on the Progression? | Rampmetrics LEAD SOURCE·FIRST AND LAST TOUCH·MULTI-TOUCH·MULTI-TOUCH WITH PATTERN RECOGNITION·MMM AND EXPERIMENTAL·LEAD SOURCE·FIRST AND LAST TOUCH·MULTI-TOUCH·MULTI-TOUCH WITH PATTERN RECOGNITION·MMM AND EXPERIMENTAL· A self-placement Most B2B marketers know exactly where they are on this journey. We're going to name the stages plainly and let you place yourself. No quiz required. # Where are you on the Progression? Most teams in B2B marketing measurement are doing one of five things. They've been doing it for a while. They mostly know what's next. They haven't moved because the next step looked underwhelming or scary or both — until recently. This document names the five stages plainly, in the language people actually use, and helps you place yourself honestly. There's no quiz. There's just a list. You'll know which one you are. A note on the framing This isn't a maturity model. We're not ranking teams. A team operating well at Stage 2 might be doing more useful work for their company than a team running fancy Stage 5 experiments. The Progression isn't a ladder you have to climb — it's a map of where the field is, and where you sit on it today. The most interesting thing about the Progression isn't the stages themselves. It's why most teams are stuck where they're stuck. The technology has been ahead of most teams for years. The reasons for the lag are mostly not technical. We'll get to that. 01 The five stages, plainly ## What people are actually doing. Read the list. You'll recognize yourself before you finish. 01 ~1975 Lead Source "We just track where the lead came from in the CRM. Usually whoever delivered the data — ZoomInfo, the form fill, the trade show, the SDR. We don't really know what marketing did before that, and we don't really track it. The slide that goes to the board uses lead source. Sometimes it's the slide that goes everywhere." This is where a startling number of B2B teams still are in 2026, especially the ones whose primary measurement is whatever shows up cleanly in Salesforce. It's not because they don't know better. It's because nothing else has felt worth the trouble. 02 MOST TEAMS First and Last Touch "We know the first touch. We know the last touch. We don't have a lot of detail on what happened between them. We've been talking about going multi-touch for years. We haven't. Honestly, we've been an ostrich about it — head in the sand, doing first and last, knowing we should be doing more, not doing more." This is the modal B2B team in 2026. Not stuck because the technology isn't there. Stuck because the upgrade has looked underwhelming for a long time — more data, more reports, more dashboards, no actual clarity. And stuck for another reason we'll name in a minute. Why so many teams stay at Stage 2 even when they could move. Here's the part of this story that nobody writes about, and it's the reason so many teams have been at Stage 2 for years. Even when the technology to do better is in place. Even when marketing teams are actively using sophisticated multi-touch analysis for their own work — the slide that goes to the board is often still lead source. Sometimes "inbound vs. outbound." Almost always something that credits sales clearly and avoids the harder conversation about what marketing actually contributed. Why? Sales compensation is tied to credit assignment. When marketing says "we want better measurement," sales hears "we want to take credit for what closed." Most of the time, marketing isn't trying to take credit — marketing is trying to optimize. But the word "attribution" makes those two sentences sound identical, and the political cost of picking the fight is high enough that most marketing leaders quietly absorb the dumbing-down rather than escalate. So the team's actual analytical capability gets used for operational work that never crosses into the boardroom. The board keeps seeing 1975. This isn't anyone's fault. It's a system producing a perverse outcome — sales is responding rationally to its compensation structure, marketing is responding rationally to the political cost of escalating, and the company collectively absorbs a worse measurement system than it's actually capable of running. The teams that genuinely move forward on the Progression are usually the ones where marketing and sales leadership have done the harder work of agreeing, in advance, that measurement isn't credit. It's how we get better together. 03 FEW DO IT WELL Multi-Touch "We capture the full journey. We can see every touch on every deal. The reports get sliced a hundred ways. We argue about credit models — first-touch, last-touch, U-shaped, W-shaped, data-driven. Sometimes the data is useful. Sometimes it's just more noise than the previous stage." Most teams aspire here. Few operate it well. The dirty secret of Stage 3: capturing the full journey without a way to read the journey produces more data, not more insight. Which is why teams who got here often felt underwhelmed and quietly went back to operating like Stage 2. Until recently. 04 NEWLY POSSIBLE Multi-Touch with Pattern Recognition "We're doing multi-touch, plus we're reading the journey. What does the path of a closed-won account actually look like? Which campaigns are surging? Which are dropping? What's the fingerprint of a deal that's going to close versus one that isn't? We're not arguing about credit models anymore — we're using the data to make better decisions." This is the upgrade that's actually worth making. What changed isn't multi-touch — multi-touch has been around for years. What changed is the pattern recognition layer on top of it. Now multi-touch gives you the journey and a way to read it. Surge detection, success fingerprinting, velocity analysis. The data starts telling stories instead of producing reports. This is where most B2B teams should be aiming today. 05 PARALLEL DISCIPLINE MMM and Experimental "We're running incrementality tests. We're commissioning marketing mix models. We're doing geo holdouts, controlled experiments, portfolio-level reallocation studies. Usually quarterly, with data scientists or external partners. We use it to validate our biggest strategic bets." This isn't a higher level than Stage 4 — it's a different discipline that runs alongside the others. Different inputs, different cadence, different expertise. Most B2B teams aren't here and shouldn't feel bad about that. The teams that are here got here because they had the data volume, the channel diversity, and the organizational capacity to act on portfolio-level outputs — not because they're more sophisticated than everyone else. Read the full MMM companion document → The whole journey, in five lines. Most teams know exactly which one they are. The harder question is what to do about it. 02 Be honest ## Place yourself. Where is your team, really? Not where you'd like to be. Not what your tool stack could do. Where your team actually operates today, and where the slide that goes to the board lands. STAGE 01 Lead Source STAGE 02 First & Last Touch STAGE 03 Multi-Touch STAGE 04 Multi-Touch + Pattern STAGE 05 MMM / Experimental A closing thought The Progression isn't a ladder you have to climb. It's a map of where the field is. Most teams should aim for Stage 4 and operate it well. Most teams shouldn't worry about Stage 5 unless the structural conditions fit. And nearly every team is held back from where they could be — not by the technology, but by the politics of who gets credit when the engine works. We made this map because nobody else was drawing it honestly. If you found yourself somewhere on it, the next step is the same regardless of stage: get your team and your sales counterparts agreeing, in advance, that measurement is not credit — it's how the company gets better together. The Progression follows from there. — THE RAMPMETRICS TEAM · COMPANION TO THE MAIN HUB → READ THE MANIFESTOREAD THE MMM COMPANION → More in this series Field Notes — a working publication. YOU ARE HERE Where Are You on the Progression? A self-placement on the five stages most teams operate at. THE MANIFESTO The only thing dead about attribution is the name. The main argument of the publication. COMPANION Is Marketing Mix Modeling For You? An honest exploration with a self-assessment. COMPANION A Brief History of Regression in Marketing 140 years of method, 50 years of marketing application. REFERENCES Sources & References The historical record behind the publication's claims. --- ## Source: https://rampmetrics.com/field-notes/attribution/two-seats The CMO and the Marketing Team | Rampmetrics A mental model # The CMO and the marketing team. Most B2B marketing organizations have two audiences for data: the CMO, who's allocating budget and presenting upstairs, and the marketing team — led by the director — who's running the actual campaigns. They need different things. The field has been pretending they need the same thing. That's the source of a lot of the current confusion — and most of the fight between MMM and multi-touch attribution. 01 The two seats ## Different jobs. Different data. SEAT ONE The CMO. JOB Set direction. Allocate budget across major buckets. Defend marketing's investment to the CFO and the board. DECISIONS Bucket-level. "More into demand gen, less into events." Is marketing working as a whole? NEEDS Strategic views. Confidence the engine is running. A defensible story upstairs. SEAT TWO The marketing team. Led by the director. Includes demand gen, campaign managers, marketing ops, content, ABM, events. JOB Translate the CMO's direction into actual campaigns. Run the work. Allocate within the buckets. Show the CMO what's working. DECISIONS Campaign and segment-level. "Which campaigns are pulling pipeline this month?" Where to lean in next week. NEEDS Touch-level customer journey data. Segment trends. Surge detection. Answers in days, not quarters. Both audiences need data. Both audiences need different data. 02 Why this matters ## The current fight, explained. The current debate about marketing measurement — "attribution is dead, MMM is the future" versus "MMM doesn't fit B2B, use multi-touch" — makes more sense once you notice which seat each side is writing for. The MMM voices are mostly writing to the CMO. Portfolio-level outputs match the CMO's decision space. Clean knobs, big buckets, quarterly cadence — that's what an MMM produces, and it's what the CMO can take to the board. The multi-touch voices are mostly writing to the marketing team — the director and the people they lead. Touch-level data, segment trends, week-to-week operating signals — that's what the team needs to actually run the work. Each side is right about its audience. The mistake has been pretending one tool serves both — and then arguing about which one tool is the right one. There is no one tool. There are two audiences. 03 An honest acknowledgment ## When MMM actually fits. Some organizations are genuinely well-suited to MMM. Large consumer brands with hundreds of millions in spend, broad channel portfolios that include TV and out-of-home, short purchase cycles, and the organizational capacity to act on portfolio-level recommendations. For those organizations, MMM is the right answer at the CMO level, and the methods deserve the respect they get. Most B2B marketing teams don't fit that shape. Fewer channels, longer cycles, smaller volumes, smaller teams. That doesn't mean those teams are behind — it means a different kind of measurement actually fits their situation. The trouble starts when methods built for the first shape get recommended as universal. For more on this specifically, the MMM companion document has a self-assessment that helps you figure out whether your org fits the profile. 04 What good looks like ## One foundation. Two views. In B2B, you have something organizations in the first world genuinely cannot have: customer-level journey data. Every touch, every campaign, every account, traceable to specific deals. That's the foundation. From that one foundation, you can build two presentation layers: - The marketing team's view. Operational depth. Campaign performance, segment trends, surge detection, the patterns that tell the team what's working this week. Used by the director and by everyone running the work underneath them. - The CMO's view. Strategic rollups from the same data. Bucket-level trend. Confidence indicators. A view of the engine the CMO can take upstairs — grounded in what customers actually did, not in a model's estimate of what they might have done. Both views, one underlying truth. The team and the CMO are reading from the same map at different altitudes. That's the answer most B2B marketing organizations need — and it's only recently been buildable, thanks to better tooling and pattern recognition on top of touch-level data. A short note from us We build for the marketing team — the director and the people they lead — because that's where strategy meets execution, and serving them well produces views the CMO can lead with. One foundation, two views. That's the design philosophy we think fits B2B. We're holding these views loosely, as with everything else in this publication. If you see this differently, tell us. The goal is to get this right. More in this series Field Notes — a working publication. YOU ARE HERE The CMO and the Marketing Team Why two audiences need different data — and what good looks like. THE MANIFESTO The only thing dead about attribution is the name. The main argument of the publication. COMPANION Where Are You on the Progression? A self-placement on the five stages most teams operate at. COMPANION Is Marketing Mix Modeling For You? An honest exploration with a self-assessment. COMPANION A Brief History of Regression in Marketing 140 years of method, 50 years of marketing application. REFERENCES Sources & References The historical record behind the publication's claims. --- ## Source: https://rampmetrics.com/field-notes/attribution/sources Sources & References | Rampmetrics Show your work Every historical claim in this publication, with the source we relied on to make it. If we got something wrong, this is where you'd start to check. # Sources & references. We make a number of historical claims across this publication — about when regression was developed, when marketing mix modeling emerged, which companies pioneered which methods, when open-source tools were released. This page collects the sources behind those claims. We've tried to use primary sources where possible (company websites, peer-reviewed papers, official releases) and to flag the places where the historical record is approximate or contested. If you find an error, tell us. A note on the sources How we worked. For the statistical history, we leaned on academic sources — the published papers themselves where accessible, peer-reviewed historical surveys, and the encyclopedia entries that summarize them. For the marketing mix modeling history, we used a combination of academic surveys, vendor company sources (verified directly from the companies' own materials where possible), industry publications (AdExchanger, Search Engine Land, Forrester), and the technical documentation of the open-source projects themselves. A few caveats worth naming up front. The exact origin date of MMM is genuinely contested in the literature — some sources point to the 1960s, others to the 1970s or 1980s — because the technique evolved gradually from econometrics rather than being invented in one moment. We've used hedged language ("emerged in the 1960s and 1970s") where the literature is mixed. Similarly, the precise inflection points in the digital attribution era (when it started, when it began breaking down) are approximate by nature. Where a claim is well-documented across multiple sources, we cite the most authoritative one. Where a claim relies on a single source, we say so. If you find a better or contradicting source, we'd genuinely like to hear about it — this is a living document. 01 Statistical foundations ## Galton, Pearson, Fisher. Galton, F. (1886). "Regression towards mediocrity in hereditary stature." Journal of the Anthropological Institute of Great Britain and Ireland, 15, 246–263. Encyclopedia summary: en.wikipedia.org/wiki/Regression_toward_the_meanSupports the claim: Francis Galton coined the term "regression" in his 1885–1886 work on inheritance and human heights, and is credited with inventing the first form of linear regression analysis. Senn, S. (2011). "Francis Galton and regression to the mean." Significance, 8(3), 124–126. rss.onlinelibrary.wiley.comSupports the claim: Detailed account of Galton's life, the development of regression, and his other statistical contributions (correlation, quartiles, percentiles). Used to confirm dates and the broader context of his work. Aldrich, J. (2005). "Fisher and Regression." Statistical Science, 20(4), 401–417. eprints.soton.ac.ukSupports the claim: R. A. Fisher synthesized the modern regression model in 1922 by combining Gauss's least squares theory with the Pearson/Yule correlation framework. Used to confirm Fisher's 1922 contribution and the role of Pearson and Yule in the intermediate development. Stigler, S. M. (2008). "Karl Pearson's Theoretical Errors and the Advances They Inspired." Statistical Science, 23(2), 261–271. arxiv.org/pdf/0808.4032Supports the claim: Pearson's role in developing the foundations of modern regression analysis (1896 paper on correlation), including the errors that Fisher subsequently corrected in 1922. Wikipedia entry: "Karl Pearson." en.wikipedia.org/wiki/Karl_PearsonSupports the claim: Pearson's development of the chi-squared test, standard deviation, correlation, and regression coefficients during 1893–1904, and his establishment of the biometrical school of inheritance. Wikipedia entry: "Ronald Fisher." en.wikipedia.org/wiki/Ronald_FisherSupports the claim: Fisher's role in formalizing modern regression, his time at Rothamsted Experimental Station (1919–1933), the development of ANOVA (1921), and his book Statistical Methods for Research Workers (1925). "Teaching the Difficult Past of Statistics to Improve the Future." (2023). Journal of Statistics and Data Science Education. tandfonline.comSupports the claim: Honest acknowledgment of the eugenics ties of Galton, Pearson, and Fisher, and the ongoing discussions in the statistics community about how to handle their legacies. We chose to note this directly rather than gloss over it. 02 Econometrics and the marketing application ## The detour through CPG. Wikipedia entry: "Marketing mix modeling." en.wikipedia.org/wiki/Marketing_mix_modelingSupports the claim: Commercial origins of MMM, including the role of CPG companies (P&G, AT&T, Kraft, Coca-Cola, Hershey, Pepsi), Nielsen and IRI data bundling, and identification of MMA (1990) and Hudson River Group (1989) as commercial pioneers. New Path Digital. "Media Mix Modeling | A Comprehensive Guide." newpathdigital.comSupports the claim: Origins of econometric modeling in the 1930s–40s with Tinbergen and Frisch; their 1969 Nobel Prize in Economic Sciences; the 1960s–70s adoption of econometric methods by marketing scholars to study marketing-sales relationships. Arima. "The Evolution of Marketing Mix Modeling: From Slide Rules to Self-Directed Insights." arimadata.comSupports the claim: University of Chicago statisticians as developers of early formal MMM in the 1970s; the resource-intensive, months-long nature of early modeling work; and CPG industry as the primary adopter through the 1980s. Eliya. "The History of Marketing Mix Modeling." eliya.ioSupports the claim: The Four Ps framework (McCarthy and Kotler, 1960) as the conceptual foundation MMM eventually quantified; CPG companies as the early adopters of the statistical version. ScanMarQED. "Beginner's Guide to Marketing Mix Modeling." blog.scanmarqed.comSupports the claim: Kraft Foods as one of the first companies to use MMM in the 1960s; the role of Nielsen tracking data in making CPG MMM possible. Hudson River Group company website (primary source). hudsonrivergroup.com/marketing-measurementSupports the claim: Founding date of Hudson River Group (1989), one of the commercial pioneers of MMM-as-a-service. Verified directly from the company's own materials, which describe founder Sean Rice as having "pioneered the marketing mix modeling industry in the United States." M-Squared. "Marketing Mix Modeling: An Origin Story." m-squared.comSupports the claim: MMM as roughly 40 years old as a commercial discipline; CPG companies as early adopters; Coca-Cola among the very first major brands to use MMM in the 1980s. Acknowledges that "the exact origins are difficult to pinpoint" — which is why our document uses hedged language. Hungry Robot (Medium). "The Historic Evolution of Marketing Mix Modeling." medium.com/hungry-robotSupports the claim: 1980s MMM as a "meticulous, time-consuming and resource-intensive" undertaking that could take a year, accessible only to firms with vast historical data, financial resources for syndicated data, and access to econometric talent. Supports our characterization of MMM as a Fortune 500 capability for most of its history. 03 The open-source democratization ## Robyn, Meridian, PyMC-Marketing. AdExchanger. "As MMM Rides Again, Google Finds Its Place In The Conversation With Meridian." (2024). adexchanger.comSupports the claim: Meta released Robyn as an open-source MMM in 2020. Google's LightweightMMM was open-sourced in 2022. Meridian launched in March 2024 (announcement) with full release in January 2025. Industry context for the renewed interest in MMM driven by cookie deprecation and walled gardens. Forrester. "Is Google's Meridian The Right Open-Source MMM Solution For You?" (February 2025). forrester.comSupports the claim: Meridian's wide launch on January 29, 2025 via GitHub. Our characterization that open-source MMM still requires strong data engineering and data science capabilities to operate well, even though the math is freely available. MASS Analytics. "Open-Source MMM: Real Costs, Hidden Trade-Offs & How to Decide." mass-analytics.comSupports the claim: Google's official January 2025 release of Meridian; PyMC Labs's PyMC-Marketing; Uber's Orbit KTR; Meta's discontinuation of active Robyn development; Meridian and PyMC-Marketing as the two most actively maintained open-source MMM frameworks today. Eliya. "Meridian vs Robyn: A Comprehensive Comparison for 2025." eliya.io/blogSupports the claim: Technical details on Robyn (ridge regression, R-based) and Meridian (Bayesian, Python, TensorFlow Probability). Meridian as the official successor to LightweightMMM. Both being open-source but with different licensing and approaches. Linea via Medium. "Comparing Robyn vs. Meridian: What open source MMM is best for me?" (David Walsh, 2025). medium.com/@david.walsh_93068Supports the claim: Two-year history of Robyn and Meridian at the time of writing (late 2025); the three industry trends driving renewed MMM interest: attribution accuracy reduction, non-cookie-based tracking needs, and the desire to measure performance media alongside brand media. Funnel.io. "What you need to know about open-source marketing mix modeling." funnel.io/blogSupports the claim: Open-source MMM landscape in 2025; the operational reality that "open source" doesn't mean "easy to use"; the ongoing role of data integration and engineering work in successful MMM programs. Search Engine Land. "Exploring Meridian, Google's new open-source marketing mix model." (Benjamin Wenner, 2024). searchengineland.comSupports the claim: Meridian's hierarchical geo-level modeling, Bayesian methods, scenario analysis, and integration with Google Search Query Volume and YouTube reach data. Comparison with Robyn's feature set. Marketing Data Science (Joe Domaleski) via Medium. "An Introduction to Marketing Mix Modeling (MMM): How It Works and Why It's Making a Comeback." (March 2026). blog.marketingdatascience.aiSupports the claim: Marketing Science Institute (2023) panel identifying privacy regulation and the decline of cross-platform tracking as key drivers of renewed MMM interest. Used to support our characterization of the digital attribution era as a "detour." 04 Framing and context ## The other claims we make. Beyond the historical record, this publication makes a number of framing and analytical claims. These are our own arguments, not facts to be verified — but where they rest on factual context, here are the sources we drew on. Claim: "The digital attribution era was a detour, not a destination." This is our framing, but it draws on widely-documented industry developments: cookie deprecation (Chrome's announced phase-out, since pushed back multiple times), Apple's App Tracking Transparency (introduced 2021), GDPR (2018), and the broader privacy regulation environment. The Marketing Science Institute's 2023 panel discussion cited in the Domaleski article above identifies these as the drivers of renewed MMM interest, which supports the "detour ending" reading. Claim: "MMM was developed for CPG and is structurally less suited to B2B." This is widely acknowledged in the practitioner literature, including the Hungry Robot Medium piece (which describes MMM's CPG origins explicitly) and the Wikipedia entry on MMM (which notes the limitations for new products and unstable launch periods). The B2B challenge — longer sales cycles, smaller data volumes, narrower channel mix — follows directly from the technical conditions MMM needs, which we discuss in the MMM Companion document. Claim: "AI is making MMM increasingly accessible to teams without dedicated data scientists." This is our current best read, held loosely. The Funnel.io and MASS Analytics pieces both describe the ongoing operational challenges of open-source MMM (data prep, model specification, interpretation) that AI assistance plausibly addresses. We've avoided making strong claims about how fast this will progress because the evidence is still emerging. Claim: "Marketing and sales political dynamics keep teams stuck at lower stages of the Progression." This is our observation, drawn from working with B2B marketing teams. We don't cite an external source for it because we haven't found one that articulates the dynamic this clearly — though we'd welcome pointers if you know of relevant research on sales-marketing alignment around measurement. The closest adjacent work is in the SiriusDecisions / Forrester sales-marketing alignment literature, which we've drawn on conceptually. Claim: "The category name 'attribution' is causing real harm to working marketers." This is our argument, presented in the main hub. We don't claim it as historical fact. The supporting evidence is the recurring confusion we observe in industry discourse — the "attribution is dead" content cycle, the inconsistent definitions of attribution across vendors and analysts, and the procurement difficulties that follow from buying software described by a misleading category name. If we got something wrong ## Corrections welcome. This publication is a working document. If you find a factual error — a date that's off, a source we've miscited, a claim that's more contested than we've represented — we want to know. The voice of the publication is "we're trying to get this right and we're holding our views loosely," and that has to apply to the factual record too, not just to the argumentative claims. The reverse is also true: if you have a better source for something we've cited weakly, or a primary source for something we've cited secondarily, we'd like to upgrade. The goal is to be the publication that does the work other vendor content doesn't bother to do. More in this series Field Notes — a working publication. YOU ARE HERE Sources & References The historical record behind the publication's claims. THE MANIFESTO The only thing dead about attribution is the name. The main argument of the publication. COMPANION Where Are You on the Progression? A self-placement on the five stages most teams operate at. COMPANION Is Marketing Mix Modeling For You? An honest exploration with a self-assessment. --- ## Source: https://rampmetrics.com/value/problems/lead-source-problem Industry Problem # The Lead Source Problem There are real, understandable reasons why most B2B teams rely on a single lead source field. But understanding what that approach captures — and what it misses — is the first step toward making better decisions. Listen to narration The Setup ## One field. One value. A 25-year-old assumption. In Salesforce — and in the broader CRM ecosystem — there's a field called “Lead Source.” It's a single picklist. You get one value. That's it. Where did this deal come from? Pick one. There's also a companion field called “Primary Campaign Source,” which is supposed to capture the last meaningful marketing touch before the opportunity was created. Again — one field, one value. These fields have been around for a long time, and they've served an important purpose. When Salesforce launched in 1999, the buying process was genuinely simpler — or at least, it was reasonable to model it that way. One person finds you, one thing brings them in, one field captures it. Clean and functional. The challenge is that B2B sales don't work that way anymore. Deals happen over months or years, with buying groups of two, three, sometimes seven or more people, across dozens of touchpoints. The single-source model was built for a world that has changed significantly — and the teams using it are often the first to recognize its limitations, even if they don't always have the ability to move beyond it. The impossible choice You buy a list from ZoomInfo. Three months later, one of those contacts visits your booth at a trade show. That's the first real engagement. So what's the lead source — ZoomInfo (where you got the email) or the trade show (where you actually connected)? Anyone who's filled out this field has faced this kind of judgment call. There's no wrong answer — but there's also no right one, because the field was designed for a simpler scenario than the one you're actually in. The Mentality ## “It has to be one thing” The single-source field wouldn't matter as much if it were just a CRM artifact that nobody referenced. But it's deeply embedded in how organizations talk about marketing performance — and for understandable reasons. In pipeline reviews, in QBRs, in board decks — the question is always the same: “Where did this deal come from?” And the expected answer is one word. Google. Trade show. Outbound. Partner. The entire multi-month, multi-person buying journey gets compressed into a single label because that's what fits on a slide and that's what leadership needs to make fast decisions. And honestly, that instinct isn't wrong. When you're making resource allocation decisions — where to hire, where to invest, what to cut — you want clarity. One signal. One takeaway. That desire for simplicity is a strength in many contexts. The tension is that the buying process itself isn't that simple, and there's a cost to compressing it into a single-source view — even when that compression makes the boardroom conversation much easier. The reinforcement loop The mentality and the tool reinforce each other. People think in single sources because the CRM only allows a single source. And the CRM was designed that way because it reflected how deals were understood at the time. Neither the tool nor the thinking is “wrong” — they're just locked in a loop that can be hard to break, especially when there's no obvious alternative already in place. Two Questions, One Field ## Credit vs. optimization — they're not the same question Here's where the tension becomes clearest. “Lead source” is being asked to answer two fundamentally different questions — and both of them are legitimate. The credit question Who gets credit for this deal — marketing or sales? Was it inbound or outbound? This is a scorekeeping question. It wants one answer. It's about organizational accountability and budget justification. The optimization question Where should we invest more? What channels are influencing pipeline? What's working and what should we cut? This needs the full picture — every touchpoint, every influence, the whole journey. The credit question naturally gravitates toward a single source. It's a zero-sum game — if marketing gets credit, sales doesn't, and vice versa. It needs a winner. And there are good organizational reasons for wanting that clarity. The optimization question needs the opposite. It needs to see every touch, understand relative influence, and account for the fact that most deals are shaped by many things working together over a long period of time. There is no single winner. There's a system. The challenge is that both questions get funneled through the same single-value field — and when that happens, the credit question tends to win. It's simpler, it's what leadership asks for, and it's what the CRM supports. The optimization question — the one that would help the marketing team make better day-to-day decisions — often gets set aside. Not because anyone decided it wasn't important, but because the system doesn't make room for it. The real cost Marketers end up operating without the intelligence they need to make fine-grained optimization decisions. They know which deals marketing “sourced.” They don't know which channels influenced the buying group, how touchpoints compounded over time, or where incremental spend would have the most impact. It's not that the information isn't valued — it's that the current system doesn't produce it. The 80/20 Surrender ## Walking before you run Most marketers — especially the experienced ones — know the single-source model is incomplete. They understand multi-touch. They understand influence. They know the buying journey doesn't reduce to one field. But they also understand their organization. Maybe their boss has been clear that the reporting needs to stay simple. Maybe they've pitched a multi-touch approach before and it didn't land — not because it was wrong, but because the organization wasn't ready for it. Maybe they're trying to walk before they run — getting the basics solid before introducing more complexity. That's a perfectly reasonable approach, and in many cases it's the right call for where the team is right now. So they work within the system. They fill in the lead source field, knowing it's an oversimplification, and they focus on the things they can control. The detailed intelligence that would help them optimize — the influence data, the multi-touch view, the buying group analysis — stays out of reach. Not because they don't want it, but because the organizational reality hasn't caught up to the analytical need. The gap The people closest to the marketing work — the ones running campaigns, managing channels, optimizing spend — often know exactly what intelligence they're missing. They're making the best decisions they can with what's available. The question isn't whether they're doing good work — they are. The question is how much more effective they could be with a fuller picture of what's actually driving results. Zoom Out ## One deal. Two very different pictures. Here's what this looks like in practice. The question on the table: “Where did this deal come from?” The lead source answer Trade ShowOne field. One value. That's what goes on the board slide. The most pessimistic version? The lead source says “ZoomInfo” — meaning the credit goes to the database where you bought the email, not even to a real engagement. The account-level reality 25 touches over 2 years. 15 people in the buying group. 5 different channels. Trade Show Google Ads LinkedIn Email Event 15 people in the buying group VP Marketing 7 touchesCMO 3 touchesDir. Demand Gen 5 touchesMarketing Ops 3 touchesSDR Manager 1 touchContent Lead 1 touchCFO 1 touchRevOps 1 touch+7 others 3 touches The trade show was one of 25 touches. An important one — absolutely. But it was part of a two-year journey that included Google Ads, LinkedIn campaigns, email nurture, and multiple events. The buying group had 15 people engaging across all of those channels. Some had one touch. Some had seven. But what goes on the board slide? “Trade show.” And here's where it gets dangerous. When the lead source view says every deal comes from trade shows, the natural next question from leadership is: “Why are we spending money on Google Ads?” And if you can't show the influence data — if you can only show lead source — you don't have a good answer. Even though 70% of your closed-won deals may have Google Ads touches in the journey, that influence is invisible in the lead source view. The channel that's quietly doing critical nurture work looks like it's doing nothing. The real danger Investment decisions get made based on lead source data. Channels get cut because they don't show up as the “source” even though they're influencing a majority of deals. The single-source model doesn't just oversimplify the story — it actively distorts which investments look valuable and which ones look expendable. And there's another layer that makes this even more complex: over a one-to-two-year sales cycle, the buyers themselves change. Maybe the original contact — the one from the trade show — leaves the company six months in. They never even took a demo. Then the person who replaces them sees a Google Ad, clicks through, and submits a demo request. That's the moment that reignites the deal. But in the lead source view, it's not the “source.” It's the third or fourth touch. The thing that actually restarted the buying process gets no credit. This is the reality of long B2B sales cycles: there isn't just one ignition point. There are multiple moments of reignition — a new person joins the buying group, someone re-engages through a different channel, a piece of content lands at the right time. The journey has chapters, and each chapter might have a different catalyst. Meanwhile, other people in the buying group are quietly educating themselves — downloading white papers, reading case studies, attending webinars. That's all influencing the sale. But because none of it is a demo request or a “talk to sales” conversion, it gets treated as noise. In the single-source mentality, only the hand-raise matters. Everything that built the conviction to raise the hand is invisible. What's at stake This isn't about doing things wrong — it's about what becomes possible when you can see more. The reignition moments, the content that educated the buying group, the channel that nurtured for months before someone else got the “credit” — all of that is real and all of it matters. When it's visible, teams can make sharper decisions. When it's not, they're doing their best with an incomplete map. What This Looks Like In Practice ## When the measurement system can't see what's working Here's a scenario that plays out more often than most people realize — and it's worth walking through because it shows how the lead source model can have real consequences for real people. An organization has historically been sales-led. Most of the pipeline credit goes to the AE team. Lead source on most deals says “ZoomInfo” or “Trade Show” — because that's what was happening before marketing got involved. The culture is predisposed to believe that sales drives the business and marketing is a support function. Then the company decides to invest in marketing. They hire a CMO. They start doing digital advertising — LinkedIn, Google, maybe Facebook. Getting that budget approved wasn't easy. It took a real internal sell to convince leadership that spending $75,000 a quarter on digital ads was worth it. A few months in, leadership asks the natural question: “What are we getting for this spend?” So they pull up the pipeline report and look at lead source. ZoomInfo. Trade show. ZoomInfo. Trade show. ZoomInfo. Not a single deal with “Digital Advertising” as the lead source. From the leadership perspective, the conclusion feels obvious: the digital investment isn't working. $75,000 spent, zero deals to show for it. Time to cut the budget, maybe restructure the team. It's a data-driven decision — they looked at the numbers and made a call. But here's the thing: the digital campaigns were working. Those LinkedIn ads were reaching the buying group at target accounts. Google Ads were driving people to case studies and product pages. The digital touches were showing up throughout the journey — as the second, fifth, and tenth touchpoints on deals that eventually closed. They were nurturing, educating, and accelerating deals that lead source attributed entirely to something else. The measurement system just couldn't see it. The disconnect This isn't a story about anyone making a bad decision. Leadership looked at the data they had and made a reasonable call. The marketing team did good work that genuinely influenced pipeline. The problem is the gap between what's actually happening and what the measurement system can show. When the only lens is lead source, channels that nurture, influence, and accelerate deals are invisible — and the people and programs behind those channels become vulnerable to cuts that feel data-driven but are actually data-limited. This kind of scenario — where an investment gets killed because the measurement system can't capture its impact — happens quietly, regularly, across the industry. It's not always as dramatic as cutting a team. Sometimes it's a slow erosion: budgets get trimmed, channels get deprioritized, programs get shelved. All because the data being used to evaluate them can only see one dimension of a multi-dimensional story. The people making these calls aren't being careless. They believe they're being data-driven — and in a sense, they are. They're just working with data that's too narrow to show them what's really going on. And that's the core problem: feeling data-driven and actually being data-driven are two different things when the underlying data only captures a fraction of reality. A Different Frame ## You can answer both questions — they just need different tools The good news is that the credit question and the optimization question can both be answered well. They just can't be answered by the same mechanism — and that's okay. Lead source can keep doing what it does. It gives leadership a simple, single-source view for pipeline reviews. That's a legitimate use case, and there's no reason to rip it out. But the optimization question deserves its own system. One that captures every touchpoint across the buying group, understands influence over time, and gives the marketing team the intelligence they need to make real decisions about where to invest. That's not a replacement for lead source — it's a complement to it. The boardroom gets its clean answer. The team gets the depth they need. Both can coexist. For teams that are walking before they run — using lead source today and getting value from it — the natural next step isn't to abandon what's working. It's to layer in the fuller picture when the organization is ready for it. The contrast between the two views — the single-source answer and the account-level journey — often speaks for itself. Once people see what they've been missing, the conversation changes on its own. A Practical Tip ## Set the expectation before you take the job If you're a marketing leader and what we've described here resonates — if you've lived the lead source frustration, fought for budget you couldn't prove the ROI on, or watched good programs get cut because the measurement couldn't see their impact — then here's something worth considering. The best time to set the expectation isn't six months into the role. It's during the interview process. When you're talking to a company about a leadership role, be direct about what you'll need to be successful. You're going to need three things: budget for the marketing programs themselves, budget for the data infrastructure to measure what's working, and your compensation. That's it. Those are the three legs of the stool. The conversation might sound something like this: “I'm excited about the opportunity, and I want to make sure we're set up to move fast. I don't want to spend my first six months pulling manual reports, hunting through silos, and trying to piece together KPIs from scattered data. If you want me to come in and drive results, I'll need the tools to actually see what's working — not an analyst to build something from scratch, but a purpose-built data foundation so we can start making real decisions from day one.” That's not an unreasonable ask. It's the equivalent of a sales leader saying “I need a CRM” or an engineering leader saying “I need a cloud provider.” It's foundational infrastructure. And framing it that way — as a requirement for doing the job well, not a nice-to-have — signals to the company that you're serious about outcomes. It also tells you something important about the company. If they're looking for someone to wear thirty hats, work eighty-hour weeks, and build their entire analytics infrastructure from scratch while simultaneously running campaigns — that's a signal about how they think about marketing. A company that understands the value of data-driven marketing will see your ask as a sign of maturity. A company that doesn't might not be the right fit. The three-budget framework When you're negotiating a marketing leadership role, be clear about the three budgets you'll need: program spend (to actually run marketing), data infrastructure (to measure and optimize it), and your own compensation. If any of the three is missing, the other two become significantly less effective. Programs without data can't prove their worth. Data without programs has nothing to measure. And a leader without either is set up to struggle. ### The shift Lead source isn't the enemy — it's a starting point. For many teams, it's the foundation they've built their reporting on, and it delivers real value within its constraints. The opportunity isn't to tear it down, but to build alongside it — adding the full-journey, buying-group-level intelligence that turns good marketing intuition into precise, data-informed optimization. The teams that make this shift don't lose what they had. They gain the ability to see what they've been missing. Myth vs. Reality AI-Powered Analytics For Demand Generation ## Ready to optimize marketing? Join leading B2B marketing teams who've transformed their marketing with Rampmetrics. Book a DemoSee Pricing Rampmetrics Dashboard Live Dashboard Attribution Campaigns Accounts Reports Alerts Settings Pipeline Influenced $4.2M +23% Marketing ROI 340% +18% Deals Influenced 127 +31% Avg Touchpoints 8.4 +12% Pipeline by Channel LinkedIn Ads$1.8M Google Search$1.2M Email Nurture$720K Webinars$480K Recent Activity Deal influenced2m Acme Corp Campaign hit $500K18m LinkedIn Brand New MQL attributed1h TechFlow Inc Report generated3h Q4 Pipeline --- ## Source: https://rampmetrics.com/value/problems/mql-problem Industry Topic # Dear MQL haters. Everyone's bashing the marketing qualified lead. But the concept was never the problem — the implementation got gutted by an organizational dynamic that nobody talks about. Listen to narration The popular take ## MQLs are dead. Or so they say. Open LinkedIn on any given day and you'll find someone declaring that MQLs are a relic. A gumball machine. A vanity metric that floods sales with junk and gives marketing something to celebrate that doesn't actually matter. The criticism has become so mainstream that bashing MQLs is practically a content strategy. And some of that criticism has a legitimate kernel. If you're chasing leads that aren't a good fit — people at the wrong companies, with no buying authority, who downloaded a white paper on a whim — then yes, your MQL process is generating noise, not signal. Nobody wants that. But here's what gets lost in the bashing: the people who built lead scoring cared. They had meetings. They debated which behaviors should carry more weight. They thought carefully about target personas, account tiers, and threshold criteria. Building a good scoring model is detailed, thoughtful work — and the teams that did it well created something genuinely useful. Reducing all of that to “the gumball machine” is dismissive of real effort and real expertise. It's also worth asking who benefits from declaring MQLs dead. In many cases, the loudest voices are people selling an alternative — an account-based platform, a new methodology, a different software category. They need MQLs to be the villain so their product can be the hero. That's not objective analysis. That's competitive positioning dressed up as thought leadership. So if the concept of scoring and prioritizing inbound leads isn't inherently broken — and it isn't — then what actually went wrong? The original design ## What an MQL was supposed to be The marketing qualified lead was designed to do something simple and useful: put a post-it note on an inbound lead that says “this one is worth your attention.” That's it. It's a prioritization signal. A way for marketing to say to sales: out of all the leads that came in this week, these are the ones you should look at first. When done right, that prioritization is built on three dimensions: #### Behavior Fit What the person did. A demo request carries more weight than a white paper download. A pricing page visit means something different than a blog read. #### Person Fit Who the person is. Their job title, their role, their seniority. A VP of Marketing and an intern both fill out forms — but they're not the same signal. #### Account Fit Where the person works. Do they work at a company that fits your ICP? A lead from a target account is fundamentally different from a lead at a company you'd never sell to. When all three dimensions are in play, an MQL is a genuinely useful signal. It means: this person did something meaningful, they're the right kind of person, and they work at a company we care about. That's worth acting on. That's what scoring was designed to produce. The intent The MQL was never meant to be a declaration that someone is ready to buy. It was meant to be a prioritization layer — a way to surface the leads most likely to be worth a conversation, so sales doesn't have to manually evaluate every single inbound form fill. It's a filter, not a verdict. And the teams that built these models — who spent hours in rooms debating thresholds, testing criteria, refining the logic — were doing real, valuable work. The concept deserves more respect than it gets. The power dynamic ## How sales killed the scoring model Here's what nobody talks about when they bash MQLs. The reason the scoring model got gutted isn't a marketing problem. It's an organizational power dynamic. Sales teams are — understandably — terrified of missing opportunities. They don't want any system sitting between them and a potential deal. Even if they helped design the scoring model. Even if they agreed to the thresholds. The anxiety is always the same: what if the scoring is wrong and I miss a deal because marketing held it back? That fear is real and it's legitimate. Nobody wants to find out after the fact that a hot lead got buried because the scoring model underweighted it. And in most organizations, sales has more political power than marketing. So when sales says “don't hold anything back” — the scoring model bends. What happens next First, person fit gets stripped out. “Don't filter on job title — just send me everything, I'll decide if they're the right person.” Then account fit goes. “Don't filter on company — what if there's an opportunity at a company we hadn't considered?” Now you're left with behavior only. Someone downloaded something? Send it over. Someone visited a page? Send it over. The scoring model has been reduced to a single dimension — and the bar is on the floor. The result ## A one-dimensional MQL that everyone hates Once the scoring is stripped down to behavior only, the MQL becomes exactly what its critics describe: a firehose of undifferentiated leads. Marketing sends everything over because they've been told to. Sales gets buried in noise because there's no filtering. And then everyone complains that MQLs don't work. How MQL scoring was designed Three-dimensional scoringHigh signal Behavior Person Account What it actually became Behavior only — one dimensionLow signal But here's the part that doesn't get said out loud: sales is now doing the scoring manually. Every lead that comes over, the AE or SDR opens it up and makes a judgment call — is this the right person? Is this a good company? Is this worth my time? They're doing exactly what the scoring model was designed to do, except now they're doing it one lead at a time, in their heads, with no consistency and no scale. The scoring didn't disappear. It just moved from a system to a person. And it got worse in the process. #### What was removed Automated, consistent, three-dimensional scoring that could process every lead at scale and surface the ones most likely to convert — before a human ever touched it. #### What replaced it Manual, inconsistent, one-at-a-time evaluation by individual reps — each applying their own judgment, with their own biases, and no shared framework. The irony ## The people who broke it are the ones complaining about it This is the part that's hard to say diplomatically, but it's true: MQLs got a bad reputation largely because the people with the most power to influence the scoring model are the same people who demanded it be gutted. Sales pushed the bar to the floor, and then complained that what came over wasn't good enough. The concept gets blamed for a failure that was imposed on it. Meanwhile, the people writing LinkedIn posts about “the MQL gumball machine” are rarely acknowledging this dynamic. It's easier — and gets more engagement — to declare a concept dead than to explain the organizational politics that broke it. The nuance doesn't fit in a headline. And the marketers caught in the middle? They know the scoring should be better. They know three-dimensional scoring would produce higher-quality leads. But they also know that the last time someone tried to hold leads back for better scoring, sales escalated it, and the scoring model got stripped down again. So they send everything over, take the criticism, and move on. The cycle Sales demands a low bar. Marketing lowers the bar. Leads get noisy. Sales complains about lead quality. Industry thought leaders declare MQLs dead. Nobody addresses the power dynamic that caused the problem in the first place. Repeat. Part 1 takeaway ## The scoring model is sound. It just needs to be reclaimed. Scoring and prioritizing inbound leads is not a broken idea. It's a necessary one. The alternative — sending every lead to sales unscored and letting individual reps make quality judgments one at a time — is more expensive, less consistent, and harder to optimize. The path forward isn't to abandon MQLs. It's to reclaim the three-dimensional model and give it the organizational backing it needs to work. But the MQL debate doesn't stop at scoring. There's a broader industry conversation happening — and it's worth understanding what's actually useful in it and what's just noise. The bigger debate ## MQL vs. ICP: a false choice The industry has framed this as an either/or. You're either an MQL-driven organization (focused on individual leads) or an ICP-driven organization (focused on target accounts). Pick a side. The ICP approach — ideal customer profile — says: define your target accounts first. Use data to identify the 200, 500, 1,000 companies that are the best fit for your product. Tier them into enterprise, commercial, and SMB. Filter by company size, geography, tech stack, industry. Then focus your marketing and sales effort on those accounts, not on random inbound leads from companies you'd never sell to. It's a good idea. It's also not new. This is database marketing — a practice that's been around for decades. The only thing that's changed is the branding. “ICP” sounds more modern than “target account list,” but the underlying concept is the same: know who you're going after and focus your resources there. The critics position ICP as the antidote to MQLs. Stop chasing individual leads, they say. Focus on accounts. And they're not wrong about the direction — account-level thinking is important. But framing it as a replacement for lead scoring is where the argument falls apart. You still need to know which individuals at those accounts are engaging, what they're doing, and how to prioritize them. That's scoring. That's what MQLs were designed to do. The false choice You don't have to choose between leads and accounts. You need both. ICP tells you which companies to focus on. MQL scoring tells you which people at those companies are showing intent. One without the other is incomplete — accounts without lead-level signals are just a list, and leads without account context are the gumball machine everyone's complaining about. The bridge ## An MQA is just a collection of MQLs When you shift to an account-based approach, the terminology changes. Instead of a marketing qualified lead, you start talking about a marketing qualified account — an MQA. It sounds like a different concept. But when you look at what's actually underneath, it's the same building blocks. Each MQL has three dimensions: behavior fit, person fit, and account fit. When you have multiple MQLs at the same company, the account fit circle is shared — they all work at the same place. What varies is the behavior (what each person did) and the person fit (what role each person holds). Stack those individual MQLs together and you get an MQA — a view of collective engagement at a single account. Individual MQL — one person, three dimensions Single MQL One person scored across behavior (what they did), person (who they are), and account (where they work). Each dimension contributes to the overall score. MQA — multiple people, shared account Collection of MQLs — MQA Three people at the same account. Each has their own behavior and person fit. But the account fit is shared — they all work at the same company. The MQA is the sum of these individual signals, rolled up to the account level. Same concept, higher altitude This is why the MQL-vs-ICP debate is so frustrating. If you do MQLs right — with all three dimensions — you're already capturing the data you need to build an account-level view. The MQA isn't a competing framework. It's what happens when you aggregate well-scored MQLs at the same company. The individual scoring feeds the account-level picture. And when you see an account where multiple people have strong behavior fit, strong person fit, and the account itself is a tier-A target — that's your qualified buying group forming. You didn't need to abandon MQLs to get there. You needed to do them well and then look at them from a higher altitude. A legitimate gotcha: funnel math There is one real problem with counting MQLs at the lead level: it can inflate your funnel math. If you have 10 MQLs but 3 of them are from the same company, you really have 7 potential deals, not 10. Run that through a pipeline model — 20% hit pipeline, 20% close — and you're overestimating by a meaningful margin. This doesn't mean MQLs are broken. It means when you're doing funnel planning, you need to deduplicate at the account level. Count the MQLs for prioritization. Count the accounts for forecasting. Those are two different uses of the same data. The insight If MQL scoring had been implemented properly from the start — with behavior, person, and account dimensions intact — the gap between “MQL-driven” and “account-driven” would barely exist. An MQA is just MQLs viewed at the account level. A qualified buying group is just MQLs clustered by role at a target account. The building blocks were always the same. The industry just got distracted by the debate and forgot to look at how the pieces fit together. The real evolution ## From leads to accounts to buying groups Here's where the conversation gets genuinely useful. The best idea to come out of this whole debate isn't ICP itself — that's table stakes. It's the concept of the qualified buying group. The insight is simple but powerful: B2B decisions aren't made by individuals. They're made by groups of 3 to 5 people — sometimes more — who collectively evaluate, champion, and approve a purchase. If you're only tracking individual leads, you're seeing fragments. If you're only tracking accounts, you're seeing the container but not what's inside it. The buying group is the unit that actually matters. The old model Individual MQLs Score and prioritize individual leads. Each person is evaluated in isolation. No account context, no buying group awareness. This is where the gumball machine criticism comes from. The ICP layer Target Accounts Define your ideal customer profile. Tier your accounts. Focus resources on companies that fit. This adds the account dimension — but still doesn't tell you what's happening inside the account. The real unlock Qualified Buying Groups Track and prioritize the buying group as a unit. See which people at a target account are engaging, what roles they represent, and whether the group collectively shows enough signal to warrant sales attention. This is where individual behavior, person fit, and account fit come together. The qualified buying group takes the best of both worlds. It uses ICP to define which accounts matter. It uses behavior and person-fit scoring to track what's happening at the individual level. And then it rolls it up to the buying group — asking not “did one person do something?” but “is a group of decision-makers at a target account collectively showing intent?” That's a fundamentally better question. And it's the one the industry should be organizing around — instead of arguing about whether MQLs are dead. Qualified buying group — example Acme Corp Tier A · Enterprise · ICP match · 4 of 5 buying group roles engaged VP Marketing Demo request, 3 webinars, pricing page Active Dir. Demand Gen Case study download, 2 emails opened Engaged Marketing Ops Product page visit, white paper Engaged CFO Pricing page visit Aware CTO No engagement yet Not yet This is what it looks like when you bring it all together. You're not chasing a single lead. You're not just looking at an account name on a list. You're seeing a buying group forming — who's engaged, what they've done, which roles are still missing — and making decisions based on the collective signal, not a single form fill. Bringing it together ## Stop debating the concept. Fix the implementation. The MQL isn't dead. The ICP isn't a revolution. And the LinkedIn hot takes aren't helping. What's actually useful is much simpler than the debate makes it seem: First, score your leads properly — with all three dimensions (behavior, person, account), not just behavior. That's the MQL done right. Second, define your target accounts and tier them. That's ICP — good practice, not a paradigm shift. Third, track and prioritize buying groups, not just individuals. That's the qualified buying group — the real evolution that bridges the gap between lead-level signals and account-level strategy. None of these are competing ideas. They're layers of the same system. And the teams that treat them as complementary — instead of picking sides in an industry debate — are the ones that end up with a demand engine that actually works. Modernizing the concept ## The MQL wasn't broken. The tooling just caught up. Whether you call it MQL, MQA, qualified buying group — the label doesn't matter. The underlying principles have always been sound: score behavior, evaluate the person, consider the account. What's changed is that you're no longer limited to what you can manually build inside a marketing automation tool. And let's be fair — the MAPs were actually pretty good at giving you the scaffolding for scoring. That was one of their genuine strengths. But if you've set up scoring a few times, you know the ceiling. The rules are static. The data sources are limited to what the MAP can see. The output is a number — 25, 50, 75 — that means different things to different people. It works, but it's basic. What's happening now is that the same core principles are being executed at a level that wasn't possible before, because three things have converged: #### More Data Sources The proliferation of APIs means you can pull in signals that were never available before — product usage data, conversation intelligence from tools like Gong, website engagement, event activity, and more. The scoring inputs are richer and wider than anything a MAP could capture on its own. #### AI-Powered Analysis Instead of static scoring rules that someone manually configured, AI can analyze patterns across all of those data sources in real time — identifying which combinations of signals actually correlate with closed deals, and adjusting dynamically as patterns change. #### Narrative Intelligence Instead of a score of 25 — which means nothing to a sales rep — the system produces a human-readable explanation of why this account looks promising. A story, not a number. "Three people in the buying group are actively engaging, the VP just attended a webinar, and product usage spiked this week." This isn't a new framework replacing an old one. It's the same framework — behavior, person, account — with the execution constraints removed. You're not throwing out MQL scoring. You're doing it in a way that's no longer limited by what one person can manually configure in a MAP. More inputs, smarter analysis, and output that a sales rep can actually act on without needing to interpret a number. The predictive scoring companies were the first wave of this — applying statistical models to lead prioritization. Good idea, early execution. What's happening now takes that further, with AI that can process a much wider set of signals in real time and explain its reasoning in plain language. Companies doing this now There are innovative companies already building on this premise. TrailSpark, for example, is taking the multi-dimensional scoring model, connecting it to a much wider set of data sources — product engagement, conversation data, website activity, CRM signals — and using AI to produce real-time, narrative-driven prioritization. It's the same core idea. It's just being executed at a level that wasn't possible five years ago. ### The real question The question was never “are MQLs dead?” The question is: are you scoring leads with the right dimensions, looking at them at the account and buying group level, and using the best available tools to do the analysis? If your scoring is one-dimensional, your data sources are limited, and the output is a number nobody trusts — the problem isn't the concept. It's the implementation. The principles are sound. The execution is ready for an upgrade. The Lead Source Problem Myth vs. Reality ## Ready to optimize marketing? Join leading B2B marketing teams who've transformed their marketing with Rampmetrics. Book a DemoSee Pricing Rampmetrics Dashboard Live Dashboard Attribution Campaigns Accounts Reports Alerts Settings Pipeline Influenced $4.2M +23% Marketing ROI 340% +18% Deals Influenced 127 +31% Avg Touchpoints 8.4 +12% Pipeline by Channel LinkedIn Ads$1.8M Google Search$1.2M Email Nurture$720K Webinars$480K Recent Activity Deal influenced2m Acme Corp Campaign hit $500K18m LinkedIn Brand New MQL attributed1h TechFlow Inc Report generated3h Q4 Pipeline --- ## Source: https://rampmetrics.com/value/problems/myth-vs-reality Myth vs. Reality # Your Marketing Data Is Not a Gold Mine There's a persistent belief that the data sitting inside your marketing automation platform and CRM is a treasure trove waiting to be unlocked. You just need the right engineer, the right API call, the right dashboard. It's an understandable belief — and the people pursuing it are doing smart, resourceful work. But the reality is more complicated than it seems. Listen to narration The Myth ## There's gold in your MAP — you just haven't mined it yet The story goes like this: your marketing automation platform — Marketo, HubSpot, Pardot — is collecting a mountain of useful data. Activity logs, form fills, email engagement, page visits. It's all there. You just need someone to go in and unlock it. Wire up the APIs. Build the dashboards. Connect the dots. And then you'll finally know what's working. It's a reasonable assumption. These platforms are collecting a lot of data. And the people who go digging — the ops leads, the engineers, the data-minded marketers — are usually the most motivated, resourceful people on the team. They hit the Marketo API, the HubSpot API, the Salesforce API. They stitch sources together in a cloud warehouse. They layer on an attribution model. They're trying to answer the right questions: what's driving pipeline? What's the full buyer journey? What should we double down on? That instinct is exactly right. The challenge is what they find when they get there. What they expect to find A complete dataset that just needs better tooling and a talented engineer to surface the insights that have been hiding in plain sight. The Reality ## It's not a reporting problem. It's a data problem. Here's the thing: the data they need — complete, connected, trusted — often doesn't exist in those systems. Not because anyone did something wrong, but because these platforms weren't designed to produce it. Every artifact they dig up has a gotcha. The sources are siloed and narrow. Stitching them together doesn't create completeness. It creates complexity. Marketing automation platforms were built for campaign execution, not campaign analysis. They're great at what they do — running workflows, nurturing leads, sending emails. But the activity tracking doesn't meaningfully connect to business results like pipeline, revenue, or real attribution. Marketo gives you almost nothing analytical out of the box without manual workflow setup. HubSpot gives you some basic first and last touch. But even when you build the workflows, the tracking is shallow compared to what a purpose-built system produces. And the manual workflows you do set up — what we call tracking debt — are fragile, shallow, and lack the richness of specialized campaign tracking. Every one of them depends on someone having manually configured something correctly. That's a lot to ask of a team that's already busy running campaigns. It's not a foundation problem. It's a structural one — the tools just weren't built for this. What they actually find Scattered artifacts that look promising but fall apart under scrutiny. Siloed sources with narrow coverage. A data problem masquerading as a reporting problem. The Silo Problem ## Every platform tells its own story. Nobody tells the whole story. Someone Slacks the digital marketing lead: “How's our campaign doing?” Simple question. And the person on the other end knows exactly what to do — they've been through this a hundred times. The campaign is running on LinkedIn and Google. So they log into LinkedIn's analytics — which is its own silo, high-level, a bit fuzzy, and naturally designed to showcase the value of advertising on LinkedIn. Then they log into Google Ads — a little more detailed, with conversion tracking you can connect to Salesforce, but still its own world. Two platforms, two perspectives, two sets of numbers that don't naturally talk to each other. Now try to answer the real question: is LinkedIn the driver and Google the nurturer? Or is Google the driver and LinkedIn the nurturer? That's almost impossible to answer from inside the silos. Each platform shows you its own slice and — understandably — claims credit for the outcome. There's no multi-touch perspective. There's no shared attribution. You're doing silo-based analysis and hoping the pieces add up. And the time cost is real. Every one of these one-off analyses — logging in, pulling data, trying to reconcile — eats hours that could be spent on the optimization work these people are actually great at. The trade show version of this problem The trade show team comes back and says “we got 50 MQLs.” They're excited — and they should be. They worked the booth, had great conversations, scanned badges. But when you load those scans into the CRM, the data team finds that 35 of those people were already in the database — they'd attended webinars, clicked emails, engaged through other channels. The trade show team counted 50. The real number of net-new MQLs is 15. The other 35 were nurtured, not acquired. Both teams are right from their own vantage point. But without a system that connects the full journey, everyone's counting from their own silo, and the numbers don't reconcile. This isn't anyone's fault. It's a structural problem. LinkedIn analytics aren't bad. Google's conversion tracking isn't useless. The trade show team isn't wrong. Each source gives you a legitimate slice of the picture. But a slice is not the story. And when every platform is designed to showcase its own value, nobody is showing you how the pieces actually fit together. In Practice ## The first-touch/last-touch illusion Here's a concrete example of how this plays out. HubSpot has a genuinely useful out-of-the-box field called “first referring site.” It captures where a contact originally came from — Google, LinkedIn Ads, organic search. You flow it from HubSpot to your Salesforce leads and contacts, and now you have a real signal about what drove that person to you. That's valuable. The team sees it and celebrates: “We got a lead from LinkedIn!” And they're right to be excited — it's a data point worth having. But then you zoom out to the account level. ### What the team sees One lead. One field. “First referring site: LinkedIn Ads.” A clear, simple signal. Time to celebrate. ### What's actually happening 25 people at that account engaging across Google, LinkedIn, organic, email, and trade shows. This is engagement number 15, not number 1. The “first touch” is only first for that individual contact — not for the buying journey. The account has been engaged for months, maybe years, across dozens of touchpoints. And the Salesforce problem compounds it: the lead isn't connected to the opportunity. There's no native way to tie them together. So someone has to manually spot the connection and call it out. That takes real effort and institutional knowledge. The people doing this work — manually connecting the dots, tracking down signals in cryptic fields — are doing something genuinely valuable. But they shouldn't have to. They're spending their expertise on detective work instead of strategy. The field isn't wrong. HubSpot isn't broken. It's doing what it was designed to do. It's just not built to tell you the whole story across an account, across time, across every channel. It gives you a useful breadcrumb — but it's easy to mistake a breadcrumb for the full map. Legacy Thinking ## Lead source: a 25-year-old field running modern attribution Salesforce launched in 1999. And from the beginning, there's been one field on the lead object called “Lead Source.” One picklist. One value. No multi-touch. No time decay. No account-level view. Just “where did this lead come from?” — as if the answer is ever one thing. And yet a lot of companies — many of them sophisticated, well-run organizations — are still basing their understanding of what works and what doesn't on this single field. It's in every report. Every pipeline review. Every QBR has a slide with lead source breakdowns. It's easy to see why. It's simple, it's already wired into everything, and it gives you something — which feels better than nothing. Nobody wants to be the person who rips it out, because doing so means confronting how much of the reporting stack depends on a concept that was never designed for modern multi-channel marketing. It's become organizational infrastructure not because it's the best option, but because it's been there so long that everything is built on top of it. The pattern People naturally anchor to whatever's already there — especially when the gap between what they have and what they'd need feels too big to cross. That's not a failure of judgment. It's a very human response to a structural problem. But the longer the gap goes unaddressed, the harder it becomes to change. The Economics ## The analysis layer has never been easier. The data layer is the hard part. Here's the encouraging news. Once you have solid, trusted data, the world is your oyster. Claude Code, AI features inside your BI tool, custom dashboards, narrative generation — there are more ways to analyze, visualize, and tell stories with data than ever before. The analysis and storytelling layer is effectively solved. You can pick whatever tool you want and go. But none of that reaches its potential if the foundation underneath is incomplete. And building that foundation yourself is where the economics get challenging. To build your own marketing data infrastructure — ETL pipelines, a cloud data warehouse, an attribution model, identity resolution, conversion tracking — you're looking at a minimum of one to two dedicated people. They're probably smart, capable engineers. But are they B2B marketing analytics specialists who live and breathe data architecture, multi-touch attribution, and pipeline velocity? That's a very specific skill set, and it's rare. So the business requirements naturally fall on the executive team. The CMO or VP ends up slacking ideas, screenshots from their last company, things they saw on LinkedIn. They become the de facto product manager for an internal tool — on top of everything else they're doing. That's a lot to ask, and the result tends to be a company-specific implementation shaped by one person's experience rather than a frameworks-based, best-practices approach. ### Build it yourself 1–2 full-time people to build and maintain. Executive time spent on requirements. ETL tools, warehouse hosting, dashboards. Ongoing dev cycles. And when someone leaves, the knowledge walks out the door. Total cost: hundreds of thousands of dollars per year. ### Buy a purpose-built platform A trusted data foundation for ~$50K. Your team uses whatever reporting and AI tools they already prefer — on top of data that's already complete, connected, and maintained. No plumbing. No key-person risk. Just data that works. So go ahead — use Claude Code, build beautiful dashboards, create AI-powered workflows. That's genuinely exciting, and we want to see more of it. Just make sure you're building on top of a data foundation that's worthy of all that effort. The Bottom Line ## You have a data problem, not a reporting problem The industry has normalized building on foundations that were never designed for full-funnel marketing analytics. Lead source fields from 1999. First-touch tracking that ignores the account. Manual workflows that break silently. None of this happened because people made bad choices — it happened because the tools evolved for one purpose and got pressed into service for another. Salesforce isn't broken. HubSpot isn't broken. Marketo isn't broken. The people working with these tools are often doing remarkable things with what's available. But these platforms weren't built from the ground up to do what marketers now need them to do. They give you useful fragments — and with enough effort, those fragments can feel like the full picture. But they're not. ### What changes everything The answer isn't more dashboards, more API calls, or more engineers hunting through your MAP. The answer is starting from trusted, complete, purpose-built marketing performance data — and then flowing that data to every surface where your team already works. That's not an upgrade. That's a fundamentally different foundation. The Business Case for Demand Generation Beautiful Workflows ## Ready to optimize marketing? Join leading B2B marketing teams who've transformed their marketing with Rampmetrics. Book a DemoSee Pricing Rampmetrics Dashboard Live Dashboard Attribution Campaigns Accounts Reports Alerts Settings Pipeline Influenced $4.2M +23% Marketing ROI 340% +18% Deals Influenced 127 +31% Avg Touchpoints 8.4 +12% Pipeline by Channel LinkedIn Ads$1.8M Google Search$1.2M Email Nurture$720K Webinars$480K Recent Activity Deal influenced2m Acme Corp Campaign hit $500K18m LinkedIn Brand New MQL attributed1h TechFlow Inc Report generated3h Q4 Pipeline --- ## Source: https://rampmetrics.com/value/problems/ai-marketing-analytics AI-Powered Analytics # Dashboards Everywhere, Understanding Nowhere A typical B2B marketing team has 6–12 dashboards open on any given Monday morning. Every one answers “what happened?” None of them answer the question that actually matters: why, and what should we do about it? The Problem ## The Dashboard Paradox More dashboards create the illusion of understanding while actually fragmenting it. Each tool shows a slice of the funnel in its own language, its own time granularity, its own definition of “conversion.” The marketing team becomes fluent in switching between tabs, not in understanding their business. ### The CMO asks: “Why did pipeline drop 30% this month?” What happens next: 1 Demand gen checks Google Analytics — traffic is flat. Not the problem. 2 Email marketing checks HubSpot — open rates are up. Not the problem. 3 The SDR manager checks Salesforce — lead-to-opportunity conversion dropped. Maybe the problem. 4 Ops pulls a Marketo report — MQL volume is actually up 15%. Confusing. 5 Someone builds a spreadsheet combining data from three sources. Takes a day. 6 The answer: a partner channel that contributed 40% of qualified pipeline went dormant, and no other channel scaled to fill the gap. Total time to insight: 2–5 days. Involves 3–4 people, multiple tools, a spreadsheet, and at least one meeting where someone says “let me pull that data.” The Gap ## Why Traditional Dashboards Fall Short ### Dashboards show metrics, not relationships A dashboard can show you that MQL-to-SQL conversion dropped. It cannot tell you that the drop is because your channel mix shifted. That requires comparing conversion rates by source, weighting them by volume, and identifying which segment drove the blended rate change. That's analysis, not reporting. ### Dashboards live in silos The partner pipeline data is in Salesforce. The content marketing metrics are in HubSpot. The website data is in Google Analytics. No single dashboard spans all three with enough context to identify the channel mix shift. Integration tools connect the data but they don't interpret it. ### Dashboards answer the question you asked, not the one you should have asked When pipeline drops, the instinct is to check each stage of the funnel. But the real answer — partner concentration risk — isn't a funnel problem. It's a portfolio problem. AI doesn't follow the standard drill-down path. It queries broadly, identifies the actual variance driver, and follows that thread. ### Dashboards can't narrate Even when all the data is in one place, a chart of MQL-to-SQL conversion by source over time is not the same as understanding what it means. Narrative is how humans actually process complex information. "Your partner channel went dormant and content marketing couldn't fill the gap" is instantly understood. A multi-series line chart with 8 lead sources requires interpretation. The Root Cause ## It's a Data Problem The dashboards aren't the disease — they're the symptom. The real problem is what's underneath them. No tool — AI or otherwise — can analyze what it can't see. And B2B marketing data is uniquely hard to get right: ### Long Cycles A lead generated today might close in 6 months. Without complete tracking across that entire timeline, any analysis has gaps — and gaps produce wrong answers. ### Multi-Touch Journeys Webinar, whitepaper, email, pricing page, SDR call. If the data layer doesn't capture every touch, no tool can see the real pattern — and the story it tells will be incomplete. ### Fragmented Sources CRM, marketing automation, ad platforms, web analytics, event tools — each holds a piece. The data needs to be unified into one trusted dataset before anything meaningful can be found. ### Signal vs. Noise Impressions and open rates are easy to measure but poorly correlated with revenue. Trusted data means the right metrics — the ones that reveal real patterns, not just confirm vanity numbers. This is why adding another dashboard doesn't help. The problem isn't the interface — it's the data underneath it. Until the data is trusted, complete, and connected, no amount of reporting will produce real understanding. The Payoff ## When the Data Is Right, AI Changes Everything Same question. Same data sources. But with a trusted, complete foundation underneath — here's what AI produces: “Pipeline is down 30% month-over-month. The drop is concentrated in mid-funnel — MQL-to-SQL conversion fell from 18% to 11%. Top-of-funnel is actually healthy: website traffic is flat but email engagement is up 12% and MQL volume is up 15%. The issue is channel mix. Partner-sourced leads, which converted at 34% last quarter, dropped 62% in volume — your top partner channel went from 180 leads/month to 68. Content marketing partially offset the volume (+22%) but those leads convert at 8%, pulling the blended rate down. Net: you don't have a demand problem. You have a partner pipeline concentration risk that was masked by aggregate MQL numbers.” One response. Full picture. The CMO has the answer before the Monday meeting ends. Solid foundation, powerful AI When every touch is captured, every source is connected, and the numbers are trusted — AI transforms what's possible. It can analyze patterns across the full funnel, identify what's actually driving change, and narrate the story in language the whole team can understand and act on. The foundation is the hard part. But once it's there, AI doesn't just speed up analysis — it unlocks entirely new ways to see your business: visual stories that land in meetings, and workflows that deliver the right insight to the right person at the right time. The Value ## Two Layers of Value Layer 1 ### Faster answers to known questions “Why did pipeline drop?” — from 3 days to 3 minutes. The data already existed. The analysis just wasn't happening fast enough. This replaces the analyst-builds-a-spreadsheet workflow with an AI that queries the same sources and synthesizes the answer in real time. Layer 2 ### Answers to questions nobody was asking When analysis is expensive (takes days, requires a specialist), teams only investigate the obvious questions. When analysis is cheap (takes minutes, conversational), teams start asking questions they never would have before: - —“Which content topics correlate with faster deal velocity, not just more MQLs?” - —“Are our webinar leads actually converting, or do they just inflate our MQL count?” - —“What happened to the leads from that event 4 months ago — did any close?” - —“If we cut our lowest-performing channel entirely, what would blended conversion look like?” These questions go unasked in most organizations because the effort to answer them exceeds the perceived value. AI inverts that equation. The Shift ## What Changes BeforeAfter CMO asks "why did pipeline drop?" — gets answer in 3 daysGets answer in the same conversation Quarterly attribution analysis in a spreadsheetOn-demand, any time, any cut of the data Dashboards show what happenedAI explains why and suggests what to do Only obvious questions get investigatedCheap analysis unlocks exploratory questions Insights live in an analyst's headInsights are conversational and shareable 6 dashboards open on Monday morningOne conversation that spans all data sources ### The sequence matters. More dashboards won't fix a data problem. And AI won't fix it either — not without the right foundation. But when the data is trusted, complete, and connected, AI turns it into something dashboards never could: real understanding, delivered in language your team can act on. Data Storytelling Beautiful Workflows The Lead Source Problem For Demand Generation ## Ready to optimize marketing? Join leading B2B marketing teams who've transformed their marketing with Rampmetrics. Book a DemoSee Pricing Rampmetrics Dashboard Live Dashboard Attribution Campaigns Accounts Reports Alerts Settings Pipeline Influenced $4.2M +23% Marketing ROI 340% +18% Deals Influenced 127 +31% Avg Touchpoints 8.4 +12% Pipeline by Channel LinkedIn Ads$1.8M Google Search$1.2M Email Nurture$720K Webinars$480K Recent Activity Deal influenced2m Acme Corp Campaign hit $500K18m LinkedIn Brand New MQL attributed1h TechFlow Inc Report generated3h Q4 Pipeline