The Unified Measurement Framework: Combining MMM, MTA, and Incrementality
Stop pitting MMM, MTA, and incrementality against each other. The unified measurement framework uses causal inference to merge all three: no cookies required.
Three platforms, three answers, one budget to defend. Your GA4 export already holds the answer, and the first finding is free and never leaves your browser. Get your first finding free
By Joris van Huët, Founder & CEOUpdated 5 min read
Run the numbers for your store: the free holdout test planner.
The Unified Measurement Framework: Combining MMM, MTA, and Incrementality
The unified measurement framework isn’t a compromise. It’s the only way to measure marketing in a cookieless world. MMM, MTA, and incrementality don’t compete: they complete each other. The problem isn’t the tools. The problem is the industry’s refusal to admit that correlation-based attribution is dead.
Why Triangulation Attribution Beats Single-Method Dogma
Triangulation attribution isn’t a buzzword. It’s a necessity. Here’s why:
- MMM sees the forest, not the trees. It models aggregate sales against spend, seasonality, and macro trends. But it can’t tell you which ad creative drove the sale. Problem: It’s slow, expensive, and requires a PhD to interpret.
- MTA sees the trees, not the forest. Problem: It’s blind to offline channels and brand lift.
- Incrementality sees the truth, but not the whole truth. Holdout tests reveal causal impact, but they’re expensive, slow, and limited to one channel at a time. Problem: You can’t run a holdout on every campaign.
The unified measurement framework doesn’t average these methods. It uses causal inference to resolve their contradictions.
How Causal Inference Solves the Cookieless Measurement Challenge
Cookies didn’t die. They were murdered: by Apple, Google, and a decade of privacy regulations. The industry’s response? Panic. Then denial. Then a desperate scramble to bolt “privacy-safe” band-aids onto broken models.
No cookies. No third-party pixels. Just math.
- Why it works: Probabilistic matching fills gaps where cookies fail.
2. Merge MMM and MTA with Structural Causal Models
Structural causal models (SCMs) don’t just correlate spend and sales. They model the mechanisms driving behavior. SCMs combine:
- MMM’s macro view (spend, seasonality, promotions)
- MTA’s micro view (ad exposures, clicks, site behavior)
- Incrementality’s causal rigor (holdout data, A/B tests)
3. Automate Incrementality with Synthetic Control Groups
Holdout tests are gold, but you can’t run them on every campaign. Synthetic control groups use causal inference to simulate holdouts at scale.
- How it works: Algorithms create a “synthetic twin” of your audience, then compare behavior against exposed groups. No manual segmentation. No guesswork.
- Speed: Results in 48 hours, not 4 weeks.
The Unified Measurement Framework in Action
Step 1: Ingest All Data Sources
- First-party data: CRM, POS, loyalty programs
- Media data: Ad server logs, DSP reports, walled gardens
- Macro data: Weather, holidays, competitor spend
- Incrementality data: Holdout tests, geo-experiments
Key: No data silos. No “clean rooms.”
Step 2: Resolve Conflicts with Causal Graphs
Who’s right? None of them: until you model the causal relationships between channels.
- Causal graphs visualize how channels influence each other. Example: TV lifts search volume, which lifts paid search conversions.
Step 3: Tune for Incremental Sales, Not Attributed Revenue
Attributed revenue is a vanity metric. Incremental sales are the truth. The unified framework optimizes for the latter by:
- Eliminating double-counting: Causal inference assigns credit where it’s due.
- Isolating lift: Synthetic control groups measure true incremental impact, not just correlation.
Why the Industry Hates the Unified Measurement Framework
The unified framework threatens the status quo. Here’s why:
- It exposes the fraud of legacy attribution. If you’re still using last-click or linear attribution, you’re leaving 30-50% of sales on the table. The unified framework makes that painfully clear.
- It requires admitting failure. Most marketers would rather defend a broken model than adopt a new one. The unified framework forces accountability.
- It’s not a black box. Legacy platforms sell “AI” that’s really just regression with a fancy UI. The unified framework is a glass box: every decision is explainable, every model is auditable.
How to Implement a Unified Measurement Framework
1. Audit Your Current Attribution
- Question: Are you using last-click, linear, or time-decay attribution? If yes, you’re wrong.
- Action: Calculate the gap between attributed revenue and incremental sales.
3. Run Incrementality Tests at Scale
- Method: Use synthetic control groups to simulate holdouts for every campaign.
- Frequency: Weekly, not quarterly. Incrementality isn’t a one-time audit: it’s an ongoing process.
4. Tune for Incremental Sales
- Metric: Stop reporting attributed revenue. Start reporting incremental sales.
- Budget: Allocate spend based on causal impact, not correlation.
The Future of Measurement is Unified
The unified measurement framework isn’t a trend. It’s the future. Here’s why:
- Cookies are gone. Third-party data is dead. First-party data and causal inference are the only way forward.
- Walled gardens are crumbling. Google and Meta’s monopoly on measurement is over.
- Incrementality is the new standard. Attributed revenue is a relic. Incremental sales are the only metric that matters.
The question isn’t if you’ll adopt a unified framework. It’s when.The ones that wait will be left behind.
FAQs
What is triangulation attribution?
Triangulation attribution combines MMM, MTA, and incrementality using causal inference. No cookies required.
Why can’t I just use MMM or MTA alone?
MMM lacks granularity. MTA lacks accuracy. Incrementality is too slow. The unified framework merges all three, filling gaps with causal inference.
Sources and Further Reading
Go deeper: Incrementality testing, explained.
Sixty-second versions of these ideas: Causality Engine on YouTube Shorts.
Have an audience of ecommerce operators? Earn 20% recurring commission by recommending the €99 read.
Keep reading
Terms in this article
- Causal InferenceCausal inference is the set of methods for estimating how much an outcome changes because of a cause, rather than alongside it. It works by comparing what happened with an estimate of what would have happened without the cause.
- Causal ModelA Causal Model is a mathematical representation describing the causal relationships between variables, used to reason about and estimate intervention effects.
- Control GroupA control group is the part of an experiment that does not get the treatment being tested, so its outcomes stand in for what the treated group would have had without it. In marketing it is the audience, or the set of regions, held back from a campaign.
- IncrementalityIncrementality is the part of an outcome, such as orders or revenue, that a marketing activity caused: the extra results compared with what would have happened without it. It is a causal effect, not the number of conversions credited to a channel.
- Linear AttributionLinear Attribution assigns equal credit to every marketing touchpoint in a customer's conversion path. This model distributes value uniformly across all interactions.
- Marketing AttributionMarketing attribution assigns credit to marketing touchpoints that contribute to a conversion or sale. Causal inference enhances attribution models by identifying true cause-effect relationships.
- Marketing ROIMarketing ROI (Return on Investment) measures the return from marketing spend. It evaluates the effectiveness of marketing campaigns.
- Vanity MetricVanity Metric is a data point that appears impressive but does not measure actual business success. It lacks a clear causal link to business objectives.