There isn’t one definitive marketing mix modeling book; the closest things to a textbook are Hanssens, Parsons and Schultz’s Market Response Models and Leeflang, Wieringa, Bijmolt and Pauwels’s Modeling Markets, which teach how to model the way sales respond to marketing. Most practitioners pair one of those with a book on regression or Bayesian statistics, a book on causal inference, and a book on how advertising actually works, because an MMM draws on all four.
That’s how I’d describe my own reading, too. When I built my first marketing mix models I came at them from growth marketing rather than econometrics, and the books that helped most weren’t the ones with “marketing mix” in the title. They were the ones that explained why a regression coefficient can mean something other than what you hope it means, and what typical advertising effects look like so you can tell when a model has gone off the rails. Every book below was checked against its publisher, author site or DOI record in October 2026, and I’ve flagged which ones are free.
Is there a marketing mix modeling textbook?
Not with that title from a major academic publisher, as far as I could find. The academic field calls the subject “market response modeling” or “marketing models”, and its standard texts predate the recent wave of open-source MMM libraries. That’s less of a problem than it sounds: the maths of adstock, saturation and regression hasn’t changed much since Simon Broadbent and the econometricians of the 1970s and 80s (the history of MMM covers who did what). What has changed is the tooling and the emphasis on Bayesian methods and experiments, which you’ll get from library documentation and the statistics books further down.
If you want the current tooling in book-length form, the closest equivalents are free: Meta’s An Analyst’s Guide to MMM in the Robyn documentation, the Google Meridian docs and the PyMC-Marketing example gallery, all covered in the guide to MMM courses and training. Academic papers fill the rest, and the annotated list of MMM research papers is the companion to this page.
Books on market response and marketing mix models
Market Response Models: Econometric and Time Series Analysis by Dominique Hanssens, Leonard Parsons and Randall Schultz (Kluwer, second edition 2001; the first edition was 1990) is the classic academic treatment of what an MMM does: specifying how sales respond to advertising, price and promotion, choosing functional forms that allow for diminishing returns, modeling lagged effects and competition, and estimating it all from time series. It’s dense and its software references are dated, but the thinking about functional form and dynamics hasn’t aged. I’d read it after you’ve built a model or two, when you start wondering why your choices matter.
Modeling Markets: Analyzing Marketing Phenomena and Improving Marketing Decision Making by Peter Leeflang, Jaap Wieringa, Tammo Bijmolt and Koen Pauwels (Springer, 2015) is the more modern textbook, organized around the full process of specifying, estimating, validating and using marketing models. If you only buy one academic book on marketing models, I’d make it this one. The same four authors edited Advanced Methods for Modeling Markets (Springer, 2017), a multi-author volume that goes further into time-series, state-space and Bayesian methods, and it’s the one to reach for when a single regression stops being enough.
Empirical Generalizations about Marketing Impact, edited by Dominique Hanssens (Marketing Science Institute, 2009; second edition 2015), is a different kind of book and probably the most useful on this list for anyone who reviews models. It collects what decades of studies have found about typical advertising and price elasticities, promotion effects and so on, which gives you benchmarks for asking whether a model’s numbers are plausible. When a model says a channel returns ten times its cost, this is the book that tells you how unusual that would be. Hanssens is also behind Long Term Impact of Marketing: A Compendium (World Scientific, 2018), which gathers research on persistent effects, and co-edited the Handbook of Marketing Analytics with Natalie Mizik (Edward Elgar, 2018; a second edition is listed as forthcoming in 2026), which surveys most methods used in the field.
Books on how advertising works
The Long and the Short of It: Balancing Short and Long-Term Marketing Strategies by Les Binet and Peter Field (IPA, 2013) is the book every MMM practitioner should read, mostly because it explains what an MMM is likely to miss. Drawing on the IPA’s databank of effectiveness case studies, it argues that short-term sales activation and long-term brand building work differently and that measurement focused on short windows systematically undervalues brand advertising. I’ve written a separate guide to The Long and the Short of It that covers its arguments and excess share of voice in more detail. Byron Sharp’s How Brands Grow: What Marketers Don’t Know (Oxford University Press, 2010) is the natural companion, and its emphasis on penetration and reach is useful context for reach and frequency decisions.
Gerard Tellis’s Effective Advertising: Understanding When, How, and Why Advertising Works (SAGE, 2004) is the research-based counterpart, summarizing what academic studies say about how advertising effects build up and decay, which is exactly what adstock and saturation curves try to capture. Simon Broadbent, the advertising practitioner who coined “adstock” (his 1979 paper “One Way TV Advertisements Work” is the usual reference), also wrote Accountable Advertising: Handbook for Managers and Analysts (1997). It’s out of print, but it’s worth finding second-hand if you want to see how much of today’s MMM practice was already worked out by a practitioner a generation ago.
Statistics and Bayesian modeling books for MMM
An MMM is a regression with transformed inputs, so the single best investment for most analysts is a good regression book. Regression and Other Stories by Andrew Gelman, Jennifer Hill and Aki Vehtari (Cambridge University Press, 2020) is the one I’d recommend: practical, honest about what regression can and can’t tell you, and full of simulation-based checks that translate directly to reviewing a model. Pair it with the guides on multicollinearity and model validation, which are where MMM-specific regression problems show up.
For Bayesian MMM (Meridian, PyMC-Marketing and most commercial platforms), start with Richard McElreath’s Statistical Rethinking: A Bayesian Course with Examples in R and Stan (second edition, CRC Press, 2020). It’s the friendliest serious introduction to Bayesian modeling, it brings causal graphs in from the start, the author’s lectures are on YouTube, and the code has been ported to PyMC. Then move to Bayesian Data Analysis by Gelman, Carlin, Stern, Dunson, Vehtari and Rubin (third edition, CRC Press, 2013), the standard reference on hierarchical models, priors and model checking, which the authors make available as a free PDF for non-commercial use. The guide to Bayesian marketing mix modeling shows where these ideas land in practice.
The half of an MMM that people underrate is the baseline: trend, seasonality and everything that isn’t media. Rob Hyndman and George Athanasopoulos’s Forecasting: Principles and Practice (third edition, OTexts, 2021) is free online and covers decomposition, regression with time-series errors, residual diagnostics and time-series cross-validation, all of which matter for baseline sales and for reading residuals. Its code is in R.
Causal inference and experimentation books
If I could make every MMM analyst read one more book, it would be a causal inference book, because the most common serious error I see is treating a well-fitting model as proof of cause. Nick Huntington-Klein’s The Effect: An Introduction to Research Design and Causality (Chapman & Hall, and free online) is the gentlest place to start, built around causal diagrams and the question of what variation in your data identifies an effect. Scott Cunningham’s Causal Inference: The Mixtape (Yale University Press, 2021) covers similar ground with more econometrics and code in R and Stata; it’s free online, and a second edition is being written in public on the same site. Its chapters on difference-in-differences and synthetic control are the methods behind geo-lift testing.
Joshua Angrist and Jörn-Steffen Pischke’s Mostly Harmless Econometrics (Princeton University Press, 2009) is the classic on when regression estimates are causal, and their Mastering ‘Metrics (Princeton, 2014) is the more accessible version. Miguel Hernán and James Robins’s Causal Inference: What If (Chapman & Hall/CRC, 2020, free online) is the most rigorous of the group; its treatment of confounding and time-varying treatments is relevant because ad spend is a treatment that changes every week, often in response to sales. For confounding variables in particular, any of these will change how you choose controls.
Finally, Trustworthy Online Controlled Experiments: A Practical Guide to A/B Testing by Ron Kohavi, Diane Tang and Ya Xu (Cambridge University Press, 2020) is about online A/B tests rather than MMM, but it’s the best book I know on how experiments go wrong in practice. If you plan to use tests to calibrate a model, the discipline in this book is what makes the calibration worth anything.
All the books at a glance
| Book | Authors | Publisher, year | Free | Best for |
|---|---|---|---|---|
| Market Response Models (2nd ed.) | Hanssens, Parsons, Schultz | Kluwer, 2001 | No | Analysts, data scientists |
| Modeling Markets | Leeflang, Wieringa, Bijmolt, Pauwels | Springer, 2015 | No | Analysts, data scientists |
| Advanced Methods for Modeling Markets | Leeflang, Wieringa, Bijmolt, Pauwels (eds.) | Springer, 2017 | No | Data scientists |
| Empirical Generalizations about Marketing Impact (2nd ed.) | Hanssens (ed.) | MSI, 2015 | No | Everyone reviewing a model |
| Long Term Impact of Marketing | Hanssens | World Scientific, 2018 | No | Brand measurement |
| Handbook of Marketing Analytics | Mizik, Hanssens (eds.) | Edward Elgar, 2018 | No | Reference |
| The Long and the Short of It | Binet, Field | IPA, 2013 | No | Marketers, everyone |
| How Brands Grow | Sharp | Oxford University Press, 2010 | No | Marketers |
| Effective Advertising | Tellis | SAGE, 2004 | No | Marketers, analysts |
| Accountable Advertising | Broadbent | 1997 (out of print) | No | History, adstock |
| Regression and Other Stories | Gelman, Hill, Vehtari | Cambridge, 2020 | No | Analysts |
| Statistical Rethinking (2nd ed.) | McElreath | CRC Press, 2020 | No (free lectures) | Bayesian beginners |
| Bayesian Data Analysis (3rd ed.) | Gelman, Carlin, Stern, Dunson, Vehtari, Rubin | CRC Press, 2013 | Yes (PDF) | Bayesian MMM |
| Forecasting: Principles and Practice (3rd ed.) | Hyndman, Athanasopoulos | OTexts, 2021 | Yes | Baseline and seasonality |
| The Effect | Huntington-Klein | Chapman & Hall | Yes | Causal inference, first book |
| Causal Inference: The Mixtape | Cunningham | Yale University Press, 2021 | Yes | Geo tests, DiD |
| Mostly Harmless Econometrics | Angrist, Pischke | Princeton, 2009 | No | Econometrics |
| Mastering ‘Metrics | Angrist, Pischke | Princeton, 2014 | No | Econometrics, gentler |
| Causal Inference: What If | Hernán, Robins | Chapman & Hall/CRC, 2020 | Yes | Rigorous causal inference |
| Trustworthy Online Controlled Experiments | Kohavi, Tang, Xu | Cambridge, 2020 | No | Experiments |
If you’re building with AI assistants, the book I co-wrote with James Phoenix, Prompt Engineering for Generative AI (O’Reilly, 2024), isn’t about MMM at all, but the page on AI in marketing mix modeling shows how its principles apply to reviewing model work. For everything else, the reading path above is the order I’d follow, and if you’d rather learn by doing alongside the reading, the Vexpower MMM course and the pillar guide to what marketing mix modeling is are the places to start.
