METHODOLOGY ↓
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_mean
Supports 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.com
Supports 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.uk
Supports 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/0808.4032
Supports 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_Pearson
Supports 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_Fisher
Supports 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.com
Supports 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_modeling
Supports 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.com
Supports 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.com
Supports 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.io
Supports 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.com
Supports 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-measurement
Supports 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.com
Supports 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-robot
Supports 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.com
Supports 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.com
Supports 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.com
Supports 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/blog
Supports 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_93068
Supports 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/blog
Supports 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.com
Supports 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.ai
Supports 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.