MMM vs Incrementality: How We Use Both to Measure Marketing Effectiveness

What marketing activity caused additional business results? It sounds like a simple question, but in practice it is one of the hardest to answer. 

Attribution, marketing mix modelling (MMM) and incrementality testing each approach it from a different angle. Attribution follows individual customer journeys, MMM models the contribution of every channel over time, and incrementality tests whether activity produced outcomes that would not have happened anyway. 

In our conversations with marketing teams, the strongest measurement setups treat MMM and incrementality as complementary methods rather than competing technologies. This article explains how each one works, where each fits and why we use both.

What you’ll find:

• How MMM and incrementality work?
• MMM vs incrementality comparison
• When to use MMM and incrementality?
• Why use both MMM and incrementality
• Where multi-touch attribution fits into this

💡 Pro Tip

Ruler combines first-party data collection, multi-touch attribution and statistical modelling in one platform, so you can see marketing effectiveness at both journey level and channel level. Our marketing mix modelling acts as a strategic measurement layer, covering digital and offline activity and showing where your next pound of budget is likely to work hardest. It also gives you the wider context to place the results of any experiments you run. Book a demo to see how it works with your own data.

MMM vs incrementality: How do they work?

Before comparing the two, it helps to be clear about what each one actually does.

What is marketing mix modelling?

Marketing mix modelling uses historical time-series data to estimate how much each marketing channel contributes to a business outcome such as revenue, sales or qualified leads. 

The model looks at spend, impressions and results over a long period, usually a couple of years, and it separates the effect of marketing from everything else that moves your numbers. That includes seasonality, competitor activity, pricing and wider economic conditions.

Because MMM works at an aggregate level, it does not need to follow individual people. That means it can include digital channels such as paid search and paid social alongside offline activity such as TV, radio, print and out-of-home. For strategic budget allocation, that breadth is the point. You get one consistent view of every channel, rather than a separate story from each platform.

There is a practical reason this matters. 44% of marketers cite cross-channel journeys, both online and offline, as a challenge for effective marketing attribution. When customers move between a podcast, a review site, a phone call and a showroom visit, a model that works from the top down can describe the combined effect far more reliably than one trying to stitch every click together.

What is incrementality?

Incrementality answers a narrower and more direct question. Did this marketing activity generate additional outcomes compared with what would have happened without it?

That second half of the sentence is the counterfactual. In plain terms, it is the version of events where you did not run the campaign. You can never observe it directly, because you either ran the campaign or you did not. So incrementality testing builds a comparison instead. One group of people or places sees your marketing, and a similar group does not. 

The difference in results between them is the incremental lift. There are three common ways to set this up.

• Control groups. A randomly selected portion of your audience is not shown the ads, and you compare their behaviour with the group that was.
• Holdouts. You pause a channel or campaign for a defined audience or period, then measure what changes.
• Geographic experiments. You run activity in some regions and withhold it from comparable ones, which works well when individual-level tracking is limited.

The result is causal evidence. It tells you that a specific activity moved the outcome, and by how much.

MMM vs incrementality comparison

The table below summarises the main differences. Neither column is better, they simply answer different questions.

The pattern is straightforward. MMM gives you the wide view and incrementality gives you the sharp one.

When to use MMM and incrementality?

In our experience, the useful question is not which method to choose. It is which question you are trying to answer right now.

Where MMM is particularly useful

MMM earns its place when you need to understand the whole picture. It is the right tool when you want to

• measure many channels on a consistent basis, rather than relying on each platform’s own reporting
• understand offline and upper-funnel activity such as TV, radio, CTV, podcasts and PR, where a click rarely exists
• decide how to allocate budget across channels for the year ahead
• see diminishing returns, meaning the point at which extra spend in a channel produces less and less additional revenue
• forecast likely outcomes from different budget scenarios before committing spend
• keep measuring when user-level tracking is incomplete or unavailable, for example because of privacy changes or customers who convert by phone

Long customer journeys are a good example. A customer might see a paid social ad, spend weeks researching, read reviews, return through organic search and finally convert. Last-click reporting gives the final touchpoint all of the recognition and the early influence none. MMM, especially when combined with impression data, picks up that earlier influence.

Where experiments are particularly valuable

Incrementality testing is most valuable when you need proof about a specific decision. It works well when you want to

• test whether a particular channel, campaign or audience generates additional outcomes
• validate an assumption, such as a belief that brand search is driving most of your new enquiries
• test a budget change by increasing or reducing spend in selected areas and measuring the effect
• establish causal lift that you can show to finance or leadership

This matters more than it might first appear. 37% of respondents say their biggest marketing challenge is generating high-quality leads. If a campaign is reporting plenty of leads but those people would have enquired anyway, the real contribution is smaller than it looks. An experiment shows you exactly how much of it is genuinely additional.

Why we use both MMM and incrementality

This is the heart of how we think about measurement. MMM provides breadth, and incrementality provides causal validation. Each one covers the blind spots of the other.

MMM can tell you that, across the last two years, a channel has contributed a certain share of revenue, and that returns start to flatten beyond a certain level of spend. What it cannot do on its own is prove cause and effect. It works from patterns in historical data, and patterns can occasionally mislead. If two channels always scale up together, for instance, the model has to work hard to separate their effects.

Incrementality testing has the opposite profile. A well-designed experiment gives you strong evidence for one channel, in one market, over one period. But it cannot tell you how that channel interacts with the rest of your mix, how results change at different spend levels, or what happens to the next twelve months of your plan.

Putting them together works in both directions.

Experiments can inform the model. When a test shows that a channel generated a particular level of lift, that result can be used to check and refine the assumptions in your MMM. If the model suggested a channel was far more effective than the experiment found, you know to revisit the model. If they agree, your confidence in the model rises considerably.

The model gives experiments context. A single test result is a snapshot. MMM places it within the full picture, showing how that channel behaves alongside every other one, across seasons and at different budget levels. It helps you decide which experiments are worth running next, and it helps you apply what you learn at scale.

What we see with the teams we work with is that this combination changes the quality of the conversation. Instead of debating whose numbers to trust, marketing and finance can look at modelled contribution, tested lift and tracked journeys together, and agree on what is really happening.

How multi-touch attribution fits into this

MMM and incrementality are two of three layers. The third is first-party tracking and attribution, and it plays a different role from either.

Attribution gives you granular, journey-level insight, fast. It shows which keywords, ads and campaigns sit on the path to a lead or sale, which is exactly what you need for day-to-day tactical decisions. At Ruler, a first-party JavaScript tag tracks calls, form submissions, live chat enquiries and other key conversions from first visit through to revenue. Enquiries and sales are matched to CRM records, so reporting reflects revenue rather than only conversion volumes. You can also compare first click, last click, linear, position-based, time decay and data-driven models side by side.

Our data-driven attribution model goes a step further by combining click-path data with impression weightings derived from marketing mix modelling. Recognition shifts away from channels that tend to be over-attributed, such as direct and brand search, and towards upper-funnel activity that influenced the decision but never earned a click. For anyone investing in CTV, display, video or out-of-home, that gives a more complete picture.

The three layers fit together like this.

• Attribution provides fast tactical signals and customer journey detail.
• MMM provides broad coverage and strategic budget insight.
• Incrementality provides evidence of causal lift.

This is also why platform reporting alone tends to fall short. Google, Meta, LinkedIn, TikTok and other platforms each report conversions within their own attribution windows, and none of them account for what the others are claiming. Add the reported conversions together and the total often exceeds your actual sales or leads. Each platform is measuring from its own perspective, which is understandable, but budget decisions need a single source of truth.

No single methodology answers every measurement question, and we would not expect one to. Attribution is quick but limited to what can be tracked. MMM is broad but works at a higher level. Incrementality is rigorous but narrow. Used together, each fills the gap left by the others.

Putting it into practice with Ruler

Ruler’s marketing mix modelling measures the impact of every marketing channel, including digital, TV, radio, print and other offline activity, using both historical performance and forward-looking budget forecasts. It accounts for seasonality, competitor activity, economic conditions and diminishing returns across more than 30 variables at once.

The budget scenario planner then turns that into decisions. Diminishing return curves show where each channel is approaching saturation and where extra investment is likely to generate incremental revenue. You can model efficiency, growth or custom budget scenarios before committing spend, so plans rest on predicted business outcomes rather than assumptions. And because Ruler ingests data from your advertising platforms, CRM and other business systems, then pushes enriched attribution data back out, offline conversion and revenue signals can flow back into your ad platforms to improve bidding and audience optimisation.

If you run incrementality experiments, whether through your ad platforms or with a specialist partner, their results sit comfortably alongside this. They give you the causal evidence, and the modelling and first-party data give you the context to act on it.

Final thoughts

MMM and incrementality are not competing approaches. MMM shows you what is contributing to performance across your whole mix, including the offline and upper-funnel activity that clicks never capture. Incrementality shows you what is genuinely generating additional results. With first-party tracking and attribution adding the day-to-day detail, you have three layers that support one another, and a far firmer basis for deciding where to invest.

If you would like to see how this works in practice, book a demo of Ruler’s first-party tracking and marketing mix modelling capabilities.

MMM vs incrementality FAQs