Coke or Pepsi? NSYNC or Backstreet Boys? Pineapple on pizza? Some debates will live on forever. For digital marketers, the Meta attribution showdown—7-day click (7DC) or 1-day view (1DV)—has been our own version of these timeless controversies. For the uninitiated: 7DC gives credit to ads that led to conversions within seven days of someone clicking, while 1DV counts conversions within one day of someone simply viewing an ad. The controversy? 1DV often inflates numbers by crediting ads for conversions that might have happened anyway, while 7DC frequently undercounts impact by missing delayed purchase decisions.
But now, Meta has rolled out a potentially game-changing attribution setting: Incremental Attribution. Could this new approach finally end the debate that's divided marketing teams for years?
Let's dive into what it means, how it works, and whether it actually delivers on its promise of measuring what truly matters.
WHAT IS INCREMENTAL ATTRIBUTION?
Incremental attribution helps marketers identify the true impact of their ads by focusing on conversions that genuinely wouldn't have happened without ad exposure—separating customers who bought because of your ad from those who were going to buy anyway.
Meta's approach uses machine learning models powered by their repository of lift study data to predict whether a conversion would have happened anyway or was actually caused by seeing an ad. This is Meta's attempt to answer the age-old question: "Did my ad actually work, or would that customer have bought anyway?"
This feature represents Meta's most significant attribution advancement since... well, since they've been battling iOS 14.5 privacy changes.
THE PROS: WHY MARKETERS ARE GETTING EXCITED
One-Click Setup That Actually Works: Incremental Attribution lives directly in Ads Manager with a genuinely simple setup.
Real-Time Incrementality Insights: Instead of waiting weeks for lift study results, advertisers get actionable incrementality data as campaigns run.
Impressive Early Results: The beta results are pretty compelling. Boots reported a 15x increase in sales! Even accounting for some beta-program selection bias, these numbers suggest significance.
Value Optimization Compatibility: Meta's Value optimization now works with Incremental Attribution, meaning brands can optimize for high-value incremental conversions, not just any conversions. For ecommerce brands especially, this could be huge.
The Attribution Debate Solver: Instead of endless debates about 1DV versus 7DC, Incremental Attribution could provide cleaner data to determine what actually works for your specific brand and audience.
THE CONS: LET'S KEEP IT REAL
True incrementality, as our SVP of Analytics, Zafreen Zerilli, reminds us, isolates the impact of a measured tactic. Whereas a true incrementality study typically relies on holdout testing—freezing media in a control and exposed study—Meta’s solution is still modeled data. The reason Stella clients love traditional incrementality is because it is a non-controversial solution, one that doesn’t rely on models at all. A 100% verified incrementality test answers, “what are my sales when this tactic is turned on versus turned off, holding other factors constant?"—and delivers a clear understanding of business drivers. Meta tests do not hold other factors constant, as Meta doesn’t have visibility into those other factors that are driving sales, like other media channels (Google, TikTok, etc.) Accordingly, it’s difficult to assess Meta’s true incrementality. As an example, if YouTube spend went up significantly during a measured period, Meta may be stealing some of that incremental credit.
Though Meta’s incrementality solution offers sound methodology, they are essentially still grading their own homework.
Limited Availability and Support: Currently rolling out in closed beta with limited campaign support—website conversions only, specific conversion goals, and brands can't just flip a switch and turn it on for all campaigns.
Learning Curve for Performance Teams: If your team is used to standard attribution reporting, there will be an adjustment period. Incremental reporting will likely show different (often lower) conversion numbers than rules-based attribution, which can concern stakeholders.
Not a Complete Solution: This isn't a replacement for comprehensive Conversion Lift studies or Geo Holdout tests when you need to measure overall media effectiveness. It's more like attribution with training wheels.
DEEP DIVE: WHY INCREMENTALITY MATTERS MORE THAN EVER
Digital marketers often confuse correlation with causation, but just because a user saw a brand ad and then converted does not mean that ad caused the conversion.
Traditional last-click attribution is particularly useless in today's multi-touch customer journey. Someone might see your brand’s TikTok ad, research on Google, check reviews on your website, abandon their cart, see a retargeting ad on Instagram, and finally convert after getting an email. Which touchpoint gets credit? All of them? None of them?
Meta's approach leverages machine learning to predict incrementality based on patterns from thousands of lift studies in their data repository. While it's not perfect (it's still modeled data, not controlled experiments), it's a strong step toward understanding causation rather than just correlation.
IMPLEMENTATION: WHAT TO EXPECT
Setting up Incremental Attribution is refreshingly straightforward. In Ads Manager, you'll find it as a new attribution setting option. Select it, and Meta's algorithms start optimizing for incremental conversions rather than all conversions.
The key is setting proper expectations with stakeholders. Post-implementation, conversion numbers will likely look different because you're now measuring incremental impact rather than all attributed conversions. This isn't a bug—it's the feature.
Timeline-wise, expect a learning period as Meta's algorithms adjust to optimizing for incrementality rather than volume. Early beta testers report seeing meaningful improvements within 2-4 weeks.
LOOKING AHEAD: THE FUTURE OF ATTRIBUTION
Meta's move toward incremental attribution signals a broader industry shift. Google will almost certainly follow suit—they're already heavily invested in AI-driven campaign optimization with Performance Max—and they can't afford to let Meta own the incrementality conversation. Similarly, we think you can expect a TikTok incremental attribution announcement within the year.
THE BOTTOM LINE
Meta's Incremental Attribution isn't perfect, but it's the best attribution-within-platform optimization option they have released to date. It won't solve all your measurement challenges, and it's not truly incremental in the academic sense, but it's a meaningful step toward measuring what actually matters.
Looking to navigate the complex media landscape and understand what’s powering your brand—be it Meta or something else? Connect with us.
COMMENTS