INCREMENTALITY IN FOCUS: WHAT ACTUALLY DRIVES GROWTH IN PERFORMANCE MARKETING

MARKETING REPORTS ARE FULL OF METRICS THAT FLATTER, NOT INFORM. ATTRIBUTION MAKES EVERY CAMPAIGN LOOK LIKE A SUCCESS, BUT MOST OF THAT SUCCESS VANISHES UNDER SCRUTINY.

It’s tempting to trust what the dashboards say (or at the very least, it’s tempting enough to consider them a reliable proxy for performance without asking harder questions). After all, conversions are up, ROAS looks healthy, and every channel appears to be doing its job. But metrics like these can be deeply misleading. Strip away the glossy numbers, and you often find campaigns taking credit for outcomes they didn’t cause. Even though it appears to be the case, these campaigns simply aren’t driving action (no matter what your overworked marketing intern tells you) – they’re simply cashing in on what users are doing already, which isn’t exactly a vote of confidence.

This article unpacks why most performance metrics are misleading, how incrementality clears the fog, and what it takes to run campaigns that actually contribute to growth.

ATTRIBUTION IS LYING TO YOU (KIND OF)

Attribution is built to make marketing look good. It hands out credit to whatever touchpoint was easiest to measure, not necessarily the one that made a difference. That’s a problem when your budget decisions are based on metrics that look and sound great on paper, but which collapse under even a modest attempt to prove actual impact. A campaign might have been served before a user converted, but that doesn’t prove it influenced the decision. Attribution rarely separates what merely appeared from what actually persuaded.

This is how budgets bleed into campaigns that look active but do nothing. The more often something shows up near a conversion, the more likely it is to get unearned credit, creating a feedback loop where noise gets mistaken for signal. In his MarTech article How Attribution Masks What’s Actually Driving Growth, Tom Leonard captures this perfectly, explaining how attribution often rewards what’s measurable, not what works. Incrementality cuts through that illusion by asking one question: did this do anything real?

The closer an ad appears to the point of conversion, the more likely it is to claim credit, whether or not it had anything to do with the outcome. This makes certain strategies look more effective than they really are, especially when the surrounding context gets ignored. Branded search, retargeting, and multi-touch campaigns all play a legitimate role in performance marketing, but the numbers can exaggerate their influence. 

The strategies outlined below are important in their own right, but are often subject to overinvestment due to overzealous marketers taking the attributed conversions as gospel. 

BRANDED SEARCH

Bidding on branded keywords can ensure your brand stays at the top of the search engine results page, and is a great way to prevent competitors hijacking clicks destined for your website. That being said, users searching for your brand are already high intent, and many of them will convert without additional encouragement. This is especially true when you’re already ranking highly for keywords organically. So, be sure to use branded keywords sparingly! 

RETARGETING

Retargeting is an essential part of most successful campaigns, however, it’s easy to overinvest if you don’t assign budget responsibility. Many marketers will see a high ROAS and think they can scale by throwing more money at it. However, retargeting audiences are a finite resource and if you don’t invest in new audiences, then the river will soon run dry. Furthermore, each retargeting impression has a diminishing impact, so the more you spend, the less each impression could matter.

MULTI-TOUCH CAMPAIGNS

Many businesses run campaigns across a number of different platforms. A user might interact with each of these platforms along their journey to conversion, but that doesn’t mean each platform played an equal part in convincing the user to convert. In many cases, platforms can just be ‘getting in on the act’, and actually contribute very little. These situations can lead to non-incremental conversions being reported and wasted ad spend.

HOW TO ACTUALLY MEASURE INCREMENTALITY

So, what does all of this actually mean? Is attribution good or not? The answer is: it’s useful for understanding what happened, but not why it happened, or whether your marketing had anything to do with it. Attribution gives you a snapshot; incrementality gives you a diagnosis. Without that distinction, you’re just rewarding activity, not impact.

To get to the “what” and the “why”, there are four proven ways to test whether your campaigns are driving real impact or just soaking up credit by default.

HOLDOUT TESTING

This method splits your audience into two random groups: one sees the campaign, the other sees nothing. By comparing the behaviour of both, you get a clear view of what changed because of the campaign. It is widely considered the most reliable way to measure true lift.

GEO TESTING

Rather than splitting individuals, you divide entire regions and expose only some to your campaign. By comparing outcomes across similar markets, you can isolate the effect without needing user-level segmentation. This is especially useful for brands with large geographic footprints or offline sales components.

TIME-SERIES ANALYSIS

This method tracks performance trends before, during, and after a campaign to spot any meaningful shifts linked to your activity. You will need to adjust for seasonality and external factors to avoid drawing the wrong conclusions. In another fantastic blog from MarTech, titled Measuring Marketing Incrementality, the author breaks down how unreliable results can be when there’s no clear distinction between cause and coincidence. Without a proper framework, you’re left guessing whether a campaign actually influenced behaviour or just happened to be nearby when something happened. The blog makes it clear: if you don’t isolate real impact, you risk making decisions based on noise instead of substance.

PLATFORM-BASED LIFT TESTING

Platforms like Meta and Google offer built-in incrementality experiments that automate test and control splits. These tools can be useful, but they should be approached with caution. Validation and proper setup are key to ensuring the insights you get are credible and not distorted by the platform’s own reporting logic.

STOP MEASURING, START STEERING

Tracking incrementality is only useful if you’re doing something with it. Too many teams watch the numbers roll in without making the shifts that actually change them. The point of measurement isn’t to admire the data; it’s to act on it. If you’re not using what you learn to rethink targeting, refine creative, and sharpen execution, then you’re just collecting numbers like some people collect Pokémon cards (it’s meant to be an actual game, but a lot of people just buy the cards with the intent of putting them away and maybe selling them again at a future point). 

USE LIVE DATA TO REALLOCATE SPEND MID-CAMPAIGN

If you wait for the wrap-up report, it’s too late. Watch for what’s working and shift spend while it still matters. Brands that act on early signals tend to avoid waste and pick up traction where others stall. According to Google, real-time tracking enables quicker responses to performance insights, which is especially useful for high-intent seasonal campaigns.

PRIORITISE COLD AUDIENCES FOR TRUE INCREMENTAL LIFT

Retargeting loyal users might boost your numbers on paper, but it rarely grows the business. Incrementality comes from changing minds, not reassuring the ones already made up. Spend more time building campaigns for colder audiences. That’s where the real gains are.

THE PROOF IS IN THE PUDDING

Most marketing teams don’t have a performance problem. They have a measurement problem. Attribution creates the illusion of progress, while incrementality asks for proof. Once you make that distinction, the whole strategy shifts. You stop rewarding noise and start demanding evidence.

If you’re ready to build campaigns that generate genuine impact, it’s time to ditch your old agency and get some new blood in. And by that we mean you should contact us to help you build strategies that prioritise real growth, backed by meaningful measurement.