By Brian Quinn, President and GM of AppsFlyer North America
Last-click attribution has long been a staple in marketers’ measurement toolkits. It’s fast, widely adopted, and provides clear signals that can guide real-time decision-making. But while useful and widespread, it doesn’t capture the full picture — especially in an environment where conversions stem from increasingly complex, cross-channel journeys.
Last-click attribution has endured because it’s simple, familiar, and operationally useful. For fast-moving teams, it provides consistent, real-time feedback that supports tactical decisions, even if it doesn’t always reflect how people actually make choices. It’s not hard to understand why we’ve seen incrementality become the adtech buzzword of 2025. Attribution methods like last-click weren’t designed to explain why a campaign worked, just that it was working. It was designed to tell you what happened and when, but with budgets under scrutiny it isn’t enough to be the sole basis for optimization or budget allocation.
With so much nuance left on the cutting room floor, incrementality acts as a layer of sanity to back up and helps to fill the gaps. Attribution shows the flow of conversions. Incrementality confirms whether those conversions were truly caused by the ads in question. Used together, they unlock what marketers really need: a measurement system that is both responsive and real.
Attribution Is Directional. Incrementality Is Diagnostic.
Marketers know attribution isn’t perfect. But perfect isn’t the goal; practical and continuous is. Attribution provides the operational backbone of performance marketing: pacing, optimization, and near real-time insights.
Where it falls short is in proving causality. It over-credits the final click. It can’t separate organic behavior from paid influence. It struggles to assign value to touchpoints that don’t result in direct clicks or conversions, especially in emerging environments like CTV or digital audio.
Incrementality fills that gap. Through structured testing – whether via geo holdouts, ghost ads, or matched-market designs – it isolates lift. It reveals whether an ad actually changed behavior, or simply tagged along for the ride. It replaces inference with evidence.
This isn’t about replacing one with the other. It’s about pairing them to build something better.
The New Standard: Attribution + Lift
The most sophisticated marketers are no longer choosing between attribution and incrementality. They’re running both, continuously and intentionally, as part of a closed-loop learning system. We’re seeing this in action with some of the largest brands in the world.
Here’s how that looks in practice:
- Attribution remains always-on, powering day-to-day decisions and tactical performance management.
- Incrementality testing runs on a cadence — monthly or quarterly — at the channel, tactic, or campaign level.
- Lift results are used to calibrate attribution data, applying coefficients that more accurately reflect the true value of each media investment.
- Strategic budget shifts are guided by validated impact, not just observed volume.
This pairing doesn’t slow marketers down, it sharpens their instincts. Attribution tells you what’s happening. Incrementality tells you what matters.
We’re operating in a measurement environment that’s never been more complex, and dollars have rarely been more scrutinized. Privacy frameworks have curtailed cross-channel tracking. Platforms increasingly mark their own homework. And performance pressure is higher than ever.
In this context, the limitations of attribution become more acute. Upper-funnel investments are harder to justify with last-click models. Conversion credit skews toward channels with the strongest tracking signals, not necessarily the strongest influence. And surface-level ROI can diverge from true business impact.
Incrementality brings marketers back to ground truth. It’s slower than attribution, yes – but that’s the point. It’s built for validation, not velocity. And when used to recalibrate attribution models, it helps ensure that what we optimize toward is actually worth optimizing.
A Role for Every Channel — and a Clearer Picture of Value
Multi-touch environments will find the value more than others. CTV impressions, audio spots, and DOOH rarely get the last click. But their contribution to brand discovery, affinity, and consideration is real, and often measurable through lift.
With incrementality in place, these channels no longer have to fight attribution bias to prove their worth. They can be judged by their actual effect on business outcomes, not just their ability to close.
And with calibrated attribution as a constant companion, those effects can be integrated into daily workflows, turning long-term confidence into short-term action.
Outlasting the Last Click
Marketers don’t need to abandon attribution. They need to upgrade how they use it.
Attribution on its own offers convenience. Incrementality on its own offers clarity. But in tandem, they create a feedback loop that’s both fast and grounded in reality.
This isn’t a philosophical debate. It’s a practical framework. The teams that embrace both methods (and know when to apply each) will build smarter media plans, defend their investments more credibly, and ultimately drive more sustainable performance.
Because in this market, it’s not enough to know what worked. You need to know what worked for real, and why.
