Skip to content

Incrementality vs. ROAS: Find the best strategy for your campaign

We've explained what incrementality is, how to measure it, and how to ensure statistically significant results. But will it work for your campaign? Let's dive into how incrementality compares to attributed ROAS when evaluating campaign performance.


As we’ve covered, incrementality is one of the most efficient and comprehensive ways to measure campaign success. However, many marketers can fall into a trap when trying to optimize toward attributed return on ad spend (ROAS) and incrementality at the same time. Here we’ll get into the details of both strategies, and answer the question, can you optimize for both attributed ROAS and incrementality during your campaign?

To learn more about how incrementality testing works, check out our blog posts about ways to measure incrementality, and how to quantify the statistical significance of your results.

Attributed ROAS

Attributed ROAS means collecting the overall summary of events that occur within the click or view window and attributing these to the retargeting partner. During a campaign that’s measuring attributed ROAS, your retargeting partner will target those users who are most likely and very likely to complete a purchase. Your mobile measurement partner (MMP) tracks this activity, and then transmits it to Adikteev.

This method is most useful for goals that should be measured in real time, or goals that are simple to measure. It’s best for marketers that need “quick wins” or want to have something to show for their retargeting campaigns in order to increase their re-engagement efforts.


When to measure attributed ROAS

‍Attributed ROAS is useful when trying to re-engage churned users that were once high-value. Since they’ve left the app for good, they are highly unlikely to return without prompting from a retargeting campaign. In this case, 100% of the campaign performance is incremental. Once these users have been brought back to the app, you must figure out if the revenue they generate is greater than the cost to reacquire them. Essentially, ROAS must be greater than 100% after a predetermined amount of time for this to be the case. For some apps, this may be 180 days or 365 days, while others should hit 100% in a shorter time frame.

Measuring attributed ROAS is also great for apps with low conversion rates, such as subscription services. In order to achieve meaningful results, incrementality requires a high volume of conversions; there must be a minimum number of conversions in both the target and control groups. If the conversion rate is intentionally low, advertisers may not have enough users to generate statistically significant results from incrementality and should rely on attributed ROAS to determine campaign performance.

Finally, ROAS gives a more accurate picture when looking for granular data. It allows for A/B testing and determining top performing creatives. Attempting to analyze this kind of data while running incrementality will lead to inconclusive results.


‍Incrementality measures the overall impact of a campaign on individual user segments. As we explained in parts one and two, audience segments are created based on user behavior. Those segments are then shown highly personalized ads. Unlike attributed ROAS, with incrementality we’re only taking into account users who saw the ad, who didn’t see the ad, and what actions the user segments performed afterwards. In this scenario, we want to target user segments that are not highly likely to convert. This means that it might take longer to calculate and see results, but in some circumstances can offer better overall results as you’re not wasting advertising dollars on users who might convert anyway. The test is run every few weeks, but our team can run an incrementality test at any time.

When to measure incrementality

‍An incrementality test compares the behavior of two similar groups of users to determine campaign impact. The test group sees ads, and the control group does not. For most cases, this is the best way to measure the true value of your campaign. It also allows advertisers to figure out which users would have converted organically, and which users were impacted by retargeting. These insights can help advertisers optimize their campaigns and budgets in order to find the best strategy for their app.

Here’s a summary of both ideas:

Diagram comparing ROAS and Incrementality. Text: Attributed ROAS vs. Incrementality: Attributed ROAS: a small audience size and a low conversion rate. Prefered for small audiences and low conversion rate (no minimum audience size requirement). Incrementality: Bigger size audience and higher conversion rate; preferred for big audience and high enough conversion rate. Big audience size required to ensure significant results. 
Diagram comparing ROAS and Incrementality. Text: Attributed Roas is good for: churned users, everything is incremental so no users are churned; A/B tests and granular optimizations, for significance reasons you usually cannot determine the best performing creative or publisher via incrementality tests. Incrementality is good for: increasing app revenue, attributed ROAS can't tell you if users would have converted without re-engagement.

With attributed ROAS, we’re trying to get credit for as many actions as possible within the click or view-through window. We target mostly users who are already likely to convert and trying to push them to convert even more.

With incrementality, we’re playing a long game. We ignore users who are already converting in the app or likely to convert. Instead, we target users who are less likely to sign up for something, make a purchase, etc. The goal is to turn these users into repeat converters and increase their LTV.

Can you optimize for both?

‍Both attributed ROAS and incrementality have their place in a re-engagement campaign. It all depends on what your goals are. Technically, the two can be run at the same time. However, you’ll only be able to optimize for one and the results might not align if you’re trying to optimize toward both.

When it comes to optimizing and improving your retargeting campaign, you’ll have to choose one KPI and stick to it. ROAS might be low if your campaign is optimized to run incrementality. Additionally, the way an attributed ROAS campaign is set up won’t offer any insightful or valuable incrementality insights. Since both strategies require targeting different sets of users, the outcomes can't be optimized together.

Which strategy will work for your app?

‍Determining which strategy is best for your app is completely dependent on your KPIs, goals and audience size. There are a handful of situations where measuring attributed ROAS is more beneficial than incrementality. If the app already has a low conversion rate, or if you’re looking to reacquire high-value churned users, measuring ROAS will give you an accurate view of performance. However, in the majority of cases, incrementality is the preferred way to measure the success of a re-engagement campaign.

The best way to figure out what will work for your app is to consult with your re-engagement partner. As experts, they’ll be able to determine the best course of action depending on your goals. At Adikteev, we offer a complementary app audit or pre-launch analysis to define the optimal strategy for your app. Book yours today and get started.

Subscribe to our newsletter so you don't miss out on industry insights and updates!

Adikteev logo on a blue background. Text reads: Want to learn more about how incrementality can help your app? Get a custom app audit here.