Cause for growth: What incrementality reveals about marketing spend

Attribution and incrementality can give you different views of the same campaign. Attribution shows where conversions are credited, while incrementality estimates the additional impact of a campaign. Used together, they provide a more complete performance picture.

Adjust InSight adds causal context by comparing observed performance with a counterfactual: an estimate of what would likely have happened without a specific marketing action. The challenge is then knowing how to use that in your next budget decision.

An incrementality test should start with the decision you want it to inform. If you want to know whether a budget increase generated additional installs, choose installs as your target metric. If you’re interested in purchases, select the relevant purchase event instead.

This particularly matters when combining incrementality with other performance metrics. Say a campaign reports 125% attributed ROAS and a 30% incremental effect on installs. Combining those figures wouldn’t produce a more accurate ROAS, since that 30% result applies to installs. If you want to understand impact on revenue, you need to run the test using a revenue event.

Calculating incremental ROAS

To calculate incremental ROAS, run a revenue-based experiment:

Incremental ROAS = Incremental revenue ÷ Incremental ad spend

This compares the additional revenue measured by the experiment with the advertising investment behind it. Incremental ROAS doesn’t account for costs such as app store or payment processing fees, so it shouldn’t be treated as a measure of profitability.

The metric you choose determines what the experiment can tell you: an install-based experiment measures incremental acquisition, while a revenue-based experiment can support an incremental ROAS calculation. Results from one shouldn’t be used as evidence for the other.

When reviewing an InSight test, start with the incremental effect, then check the Significance shown in the Model section.

The incremental effect shows the estimated impact of your marketing action on your selected metric. Significance tells you whether there is sufficient statistical evidence behind the observed effect. So, if a large lift is marked Not significant, there isn’t enough statistical evidence to treat the lift as a reason to change your budget.

Next, look at the Organic result. This shows whether the action generated an incremental effect on organic activity or whether there is evidence of organic cannibalization. InSight brings these signals together in the overall result and recommendations shown alongside the test.

Identifying organic cannibalization

Some results indicate that marketing activity is adding less than topline performance suggests. InSight can estimate organic cannibalization, where paid activity may be replacing conversions that would otherwise have happened organically.

Say you increase a campaign's budget and paid installs rise. If InSight also identifies a decline in Organic installs associated with that action, some of the additional paid activity may be replacing demand that would have existed anyway.

Evidence of organic cannibalization can therefore be an important signal when deciding whether to maintain the investment or move budget elsewhere.

Using incrementality results to guide spend

Your InSIght results can help inform what you do with your budget next. The incremental effect, its statistical significance, and your campaign targets provide context for deciding whether the result supports a change in spend.

Increase spend: Strong positive incrementality that is statistically significant, delivers meaningful incremental volume, and exceeds your target can support additional investment.

Maintain spend: Positive incrementality with stable returns can support keeping investment at its current level.

Reduce or reallocate spend: Low or absent incremental lift, or evidence of organic cannibalization, can indicate that the budget may be more useful elsewhere.

These signals don’t determine the decision for you. Your objectives and performance targets still define what makes a result worth acting on.

Put incrementality into your measurement mix

No single measurement method gives you the full picture of campaign performance. Attribution can show where conversions are credited, while incrementality adds evidence of what your marketing action actually contributed.

That causal view becomes particularly useful when you’re deciding whether to change spend. InSight can help you test the impact of that decision, identify where paid activity may be cannibalizing Organic demand, and give you stronger evidence for what to do with your budget next.

The aim isn’t to replace the metrics you already use. It’s to add the context you need to make better decisions from them.

Request a demo** today to learn how Adjust can help you improve retention and grow your app business.**