Do we need a media mix model at our spend?
Our answer
Most businesses don't need a media mix model yet. Below about $50K a month across channels, rely on store revenue, platform reports and lift tests. Consider one at about $3M to $5M a year across three or more channels.
- Confidence
- Medium
- Strongest proof
- 1 official rule
- 3 practitioners, 9 opinions or claims
- Sources
- 12 sources
- 13 statements
New here? Take the 1-minute tour
Why this is the answer
3 reasons and our judgmentFlighted, a paid media agency, says you generally need a model once you spend $50K or more a month across channels. Eightx, a fractional CFO firm, puts the floor at $3M to $5M a year across 3 to 5 or more channels.
A model needs years of weekly data. Eightx asks for 18 to 24 months. Google's Meridian guide says two years, 104 data points, is too few to estimate 12 channels reliably, and suggests fewer or combined channels, three years or regional data.
An unchecked model can mislead. Michael Kaminsky of Recast, which sells mix models, argues most businesses do not need one. Boathouse saw an off-the-shelf model credit ads with 75% of new customers where a custom build said under 1%.
Our judgment
Below about $50K a month across channels, every source that gives a spend level says you don't need a model yet. Between that and $3M to $5M a year the sources split, so a model there fits only as a managed or pilot setup checked against a geo test. Above it they agree a model is justified once spend spans three or more channels. Treat about two years of weekly spend and revenue as the minimum. The sources ask for 12 months (WorkMagic) to three years (Google), and two years is the top of Eightx's 18 to 24 months and the bottom of Cometly's two to three years. Check any model against a holdout forecast and your own lift tests before moving budget, because the simulated 10.61 times against a true 4.20 times and Michael Kaminsky's disagreeing runs show how far an unchecked model can drift.
Our judgment, built from Flighted, Eightx, Cometly, Elevarus, WorkMagic, Google, Sonia Chung, PPC Land and Michael Kaminsky: our reading of these sources, not a finding from any one of them.
What the sources say
13 statements from 12 sourcesBoathouse's strategy chief says that for one client an off-the-shelf mix model credited paid media with about 75% of new customers, while a custom model on the same data said under 1%. Check models with lift tests.
“A result below 1% deserves as much scrutiny as a result of 75%.”
Meta's Robyn mix model docs say ad returns diminish, so budgets should follow the next dollar's response, and models should be calibrated with lift or geo tests. A relayed whitepaper puts uncalibrated models 25% off.
“uncalibrated models show 25% average difference to the ground truth”
Gruns' growth head matches the tool to the decision: incrementality tests, a mix model and marginal return for big budget calls, and attribution and platform data for daily moves.
“I'm also not letting attribution decide whether Meta deserves another $500K.”
In a simulation by a Zalando researcher, a standard mix model reported 10.61x ROAS for paid search whose true return was 4.20x; estimating from four geo experiments gave 4.14x.
“Structural estimation across four geo-experiments landed on 4.14x against a true 4.20x”
WorkMagic says a media mix model needs about 3,000 to 5,000 orders a month, 12 to 24 months of varied spend history and several channels. With Google Ads alone it won't help.
“if Google Ads is your only line item”
Michael Kaminsky, whose firm Recast sells mix models, says most businesses do not need one and it can mislead them; he cites studies where last-touch was about 85 percent as good as experiments in simpler mixes.
“most businesses do not need an MMM. In fact, an MMM will be net negative for them because it will actively mislead them.”
Flighted says a mix model fits brands spending $50K+ a month across 3 or more channels, while Meta-only brands with same-day purchases do better with incrementality tests.
“spending $50K+/month across channels”
AdExchanger reports, from vendor and agency sources, that Meta has stopped pushing its Robyn mix model while Google's sales force drives Meridian. A rival vendor's CEO says Meridian needs customizing.
“it’s free in the sense that a puppy might be free, not like a free beer”
Haus, a measurement vendor, says a media mix model fits once you run a real mix of channels, buy ones regional holdouts cannot test, and already act on experiments. It gives no spend thresholds.
“If experiments are still something you run and then quietly ignore, an MMM built on top of them won't get the trust it needs.”
Google says a mix model with 26 parameters on two years of weekly national data is too little to estimate reliably; fewer channels, more years or geo-level data help. It calls the guidance rough.
“This sample size is too low to estimate the model reliably.”
Eightx puts the mix model adoption floor near $20M+ revenue and $3M to $5M+ a year in media across 3 or more channels; under $1M, use platform reports plus quarterly incrementality tests.
“The adoption floor for DTC is roughly $20M+ in revenue and $3M-$5M+ in annual media spend”
Elevarus says most accounts need 18 months of weekly spend and revenue for a mix model, single-channel advertisers don't need one, and a small geo test should come first. No data.
“Single-channel advertisers do not need this. Multi-channel advertisers do.”
Cometly, which sells attribution, says a mix model needs two to three years of data and at least four channels; under $100,000 a year across fewer than four channels, focus on attribution first.
“The practical threshold for most small businesses: if you're spending less than $100,000 annually on marketing across fewer than four channels, focus on attribution first.”
In the full answer
For readers who sign in- What to do this week3 steps
- The numbers behind it3 figures from cited sources
- Where each source standsA chart placing each source on the question
- The strongest disagreementThe best case against the answer, from a cited source
- How sure we areWhat its confidence rests on, and what would change the answer