For performance marketers, that means being able to separate genuine campaign uplift from conversions that would have happened anyway. Organic behavior, brand equity, email, push notifications, other paid channels and external factors are already present in the control group. Incrementality measures the additional value created on top of that baseline.

What is incrementality?

Incrementality is a measurement methodology used to determine the causal impact of advertising. Users or markets are divided into comparable treatment and control groups: one group is exposed to a campaign and the other is not. The difference in their outcomes represents the campaign's incremental impact.

The methodology is based on the same principle used in randomized controlled trials. A conversion occurring after exposure to an ad does not necessarily mean the ad caused it. Incrementality is designed to answer that causal question.

Incrementality is how Remerge measures app retargeting campaigns, and the same method applies to user acquisition.

Incrementality in practice

Incrementality is only useful if the methodology stands up to scrutiny. That means transparent test design, balanced groups and results that can be independently validated.

Miniclip verified the methodology

When Miniclip ran an incrementality test with Remerge for its 8 Ball Pool retargeting campaigns, the team went beyond simply accepting the reported results. Miniclip compared Remerge's randomization data with its own internal behavioral data, including activity, matches played, player level, lifetime value and retention.

Using Distribution Matching, Miniclip found that the treatment and control groups were perfectly balanced across the KPIs it examined - an independent validation of the randomization methodology.

“Remerge's transparent approach towards incrementality testing has made it easy to verify their results from our end.”
- Jamie Lee, Digital Marketing Analyst, Miniclip

Read the Miniclip incrementality study

Socialpoint found its lowest-ever incremental acquisition costs

Socialpoint used Remerge's Causal Impact methodology to measure the incremental performance of User Acquisition campaigns for Dragon City. Over an eight-week period, the teams ran 14 tests and experimented with different creative formats.

The results delivered Socialpoint's lowest-ever iCPI and iCPA figures, while also showing that video outperformed static and native formats in both CTR and conversion rate. The tests gave Socialpoint actionable evidence for how to optimize future UA activity rather than relying solely on attributed performance.

“We are doubling down on building tools that will enable our UA teams to measure actions with 100% incrementality in the event that all other types of measurement move away.”
- Alexandre Noirot-Cosson, Head of Marketing for Live Games, Socialpoint

Read the Socialpoint case study

PhotoSì identified when to spend

For seasonal businesses, performance can change dramatically depending on the time of day, day of week and wider calendar. PhotoSì worked with Remerge on a series of incrementality tests over four weeks, testing supply partners, creatives and campaign timing across ID and no-ID traffic.

The findings showed that weekends and evenings were the strongest times to run campaigns, while video and interstitials were the best-performing creative formats. The predictive model also highlighted the importance of seasonality, helping PhotoSì make more informed decisions about where and when to allocate budget.

“Incrementality testing gives you another perspective of your User Acquisition activity, which you don’t have with a deterministic approach.”
- Manuel Mainetti, PhotoSì

Read the PhotoSì case study

Delivery Hero measured value in a no-ID world

Delivery Hero became one of the first advertisers to use Remerge's incrementality solution for User Acquisition on no-ID traffic. Starting with HungerStation in Saudi Arabia, the teams first ran multiple experiments on ID traffic using different creative formats during peak evening hours.

Those learnings were then used to inform tests on no-ID traffic. The project helped achieve competitive CPI performance, identify campaign elements for future optimization and demonstrate the incremental value of the advertising activity.

Read the Delivery Hero case study

How to run an incrementality test

A successful incrementality test starts with a business question, not a methodology. Before launching a campaign, define exactly what you need to prove.

1. Define the conversion event

Decide what outcome the campaign should create. Depending on your business, this could be an install, registration, purchase, order, re-engagement or incremental revenue. The KPI should reflect the business objective rather than simply the easiest event to measure.

2. Create comparable test and control groups

For retargeting, users should be randomly assigned so that they remain in either the treatment or control group. This is essential for avoiding overlap and ensuring that differences in outcomes can be attributed to the advertising exposure.

For no-ID campaigns, comparable treatment and control sub-markets can be used instead, allowing aggregated data and econometric analysis to measure campaign uplift.

3. Run long enough to generate meaningful data

The right test duration and group size depend on audience volume, expected conversion rates and the size of the effect being measured. A test needs enough data to distinguish a genuine signal from normal variation.

4. Compare outcomes, not just attributed conversions

The control group represents the baseline: conversions and revenue generated through organic behavior, other marketing activities and external influences. The treatment group shows what happened with the advertising campaign running.

The difference between those outcomes is the incremental contribution of the campaign.

5. Use the results to optimize decisions

The value of incrementality is not simply proving whether a campaign worked. Results can be used to compare creatives, audiences, supply partners, times of day, geographies and optimization strategies, and then move budget toward activity that creates genuine additional value.

How to read incremental ROAS

Traditional ROAS measures attributed revenue against advertising spend. Incremental ROAS, or iROAS, focuses only on the additional revenue caused by the campaign.

Incremental revenue ÷ ad spend = iROAS

For example, if an exposed group generates more revenue than a comparable control group, the difference is the campaign's incremental revenue. Dividing that additional revenue by spend gives you iROAS.

Worked examples are in our guide to interpreting incremental ROAS.

The same principle applies to other incremental KPIs:

• iCPI measures the cost of each additional install generated by the campaign.
• iCPA measures the cost of each additional conversion or action.
• Incremental uplift measures the difference in conversion or revenue outcomes between treatment and control groups.

What is a good incremental ROAS?

There is no universal benchmark. A good iROAS depends on your margins, customer lifetime value, payback period and the conversion event being optimized.

For example, an e-commerce app may be able to evaluate incremental revenue directly. A food delivery app with relatively fixed profit per order may prefer to set an iCPA target based on the profit generated by each incremental conversion.

The important distinction is between attributed performance and incremental performance. A campaign can report a strong attributed ROAS while contributing little additional revenue if many of those users would have converted anyway. Incremental KPIs provide the evidence needed to make that distinction.

Frequently asked questions about incrementality

What is incrementality in marketing?

Incrementality measures the additional outcome caused by a marketing activity. It compares the behavior of an exposed treatment group with an unexposed control group to determine whether the advertising created genuine uplift.

How is incrementality different from attribution?

Attribution assigns credit for a conversion to one or more marketing touchpoints. Incrementality measures whether a campaign actually caused an additional conversion, action or revenue outcome. The two approaches can work alongside each other: attribution can provide standardized operational reporting, while incrementality can validate whether attributed results reflect genuine value.

We compare the two approaches in more depth in Attribution vs incrementality: pros and cons.

What is incremental ROAS?

Incremental ROAS, or iROAS, is the revenue generated in addition to what would have happened without advertising, divided by ad spend. It is designed to measure the true return created by a campaign rather than all revenue attributed to it.

How does incrementality prevent double-paying for users?

Incrementality does not attempt to decide which partner should receive attribution for an individual conversion. Instead, it measures the total additional effect of a campaign against a control baseline. If users would have converted without a campaign, that behavior is reflected in the control group rather than counted as incremental value.

Can incrementality be measured without device IDs?

Yes. Causal Impact methodologies can use aggregated data and identifiable sub-markets rather than individual user IDs. This makes incrementality particularly useful for measuring the effectiveness of no-ID campaigns.

How do you know an incrementality test is reliable?

Reliable tests require comparable treatment and control groups, sufficient sample size and duration, clear KPI definitions and a methodology that can be independently verified. Transparency is critical: advertisers should be able to understand how groups were created and validate the analysis with their own data where appropriate.

Should incrementality replace traditional attribution?

Not necessarily. Attribution and incrementality answer different questions. Attribution helps marketers report and operate campaigns across channels, while incrementality helps validate whether those campaigns are creating additional business value. Used together, incrementality can provide an important calibration layer for attribution models.

When should I use incrementality testing?

Incrementality is particularly valuable when you need to prove campaign uplift to leadership, evaluate the true value of retargeting, compare partners or channels, optimize budget allocation, measure no-ID traffic or investigate whether attributed conversions are cannibalizing organic behavior.

Measure what your campaigns actually change

Clicks and attributed conversions can tell you what happened after an ad was served. Incrementality helps answer the harder question: what would have happened without it?

That difference is what performance teams can take to leadership when defending spend, evaluating partners and deciding where the next budget should go.