If you've ever looked at your Facebook ad results and wondered, "How many of these conversions would have actually happened anyway?" — you're asking exactly the right question. The answer lies in Meta's new attribution model that's changing how we measure ad performance.

In April 2025, Meta rolled out Incremental Attribution in Ads Manager — a feature that separates real ad-driven conversions from those that would have occurred naturally. Early adopters across 45 advertisers and 11 verticals saw an average improvement of more than 20% in incremental conversions.

20%+
Improvement in incremental conversions
45
Advertisers in Meta's initial test (Jan–Jun 2024)

What is Incremental Attribution?

Incremental Attribution is a new approach that helps performance marketers identify the true impact of their ads by isolating incremental conversions — purchases or actions that would not have occurred without exposure to your ad.

Unlike traditional attribution models that simply measure what happens after an ad exposure, this approach focuses on measuring the incremental lift: what your ads actually caused, versus what was going to happen anyway.

  • AI-Powered Optimisation: Uses machine learning and Meta's Lift study data to optimise campaigns toward incremental outcomes
  • Holdout Testing Methodology: A portion of the audience is withheld from seeing the ad (holdout group) while the rest are exposed (test group)
  • Real Performance Measurement: Distinguishes between organic conversions and those genuinely driven by advertising

How It Works

The Holdout Testing Methodology

The system operates through a straightforward but powerful methodology:

  1. Holdout Group: A portion of your target audience is excluded from seeing your ad.
  2. Conversion Comparison: Meta compares conversion rates between the exposed and holdout groups to determine incremental lift — the additional conversions attributable to the ad.
  3. Machine Learning Analysis: Meta employs statistical techniques including Pre-Post Analysis (Difference-in-Differences) to analyse the holdout test results at scale.

How It Differs from Traditional Attribution

Traditional models credit conversions based on clicks or views — if someone clicked your ad and then purchased, the ad gets credit, even if they were going to buy anyway.

Incremental Attribution measures the "what would have happened if the ad was never shown" part of the equation. It's the difference between correlation and causation.

Meta's Attribution Settings in 2025

Meta now offers several attribution options. Understanding when to use each is critical for accurate measurement.

E-commerce / Quick decisions

7-Day Click + 1-Day View

The default setting. Works best for products requiring a few days to decide — furniture, electronics, items with many substitutes.

High-value / B2B

7-Day Click Only

For purchases that take time and complex sales cycles. More conservative — prevents inflated numbers from view-through attribution.

Impulse / Low-cost items

1-Day Click

Best for products decided at first sight — cheap items, snacks, stationery. Reduces attribution of delayed purchases that may not be ad-influenced.

Maximum insight

Incremental Attribution

Focuses on true incremental conversions. Clearest picture of actual ad impact. Best for prospecting campaigns.

When to Use Incremental Attribution

It Works Best For:

  • Prospecting campaigns — bringing in new customers who would not have purchased otherwise
  • Budget allocation decisions — directing funds toward campaigns that genuinely drive additional conversions
  • Cross-channel analysis — avoiding giving Meta credit for conversions that would have happened from another channel

It's Less Ideal For:

  • Retargeting campaigns — these target users who have already shown interest, so many would convert anyway, making incremental value appear lower
  • Sales periods — during promotions, use 1-day click + 1-day view to let Meta optimise as fast as possible to capture customers before the sale ends

How to Set Up Incremental Attribution

  1. In campaign setup within Ads Manager, locate "Show more options" under your performance goal section
  2. Select Incremental Attribution as your attribution setting
  3. Let Meta's AI-driven model run — the system will automatically handle the holdout groups and measurement
  4. Once live, Incremental Attribution results appear directly in Ads Manager — initially only in the results column

You can also add the 'Incremental attribution' column to your reporting even if you're not optimising for it — useful for benchmarking your existing campaigns against incremental outcomes.

Best Practices

  • Don't change what's working. If a campaign is performing very well with current attribution settings, leave it alone. Test incremental attribution on new campaigns first.
  • Expect lower conversion numbers. This isn't a problem — it's accuracy. Incremental attribution shows reality, not inflated numbers.
  • Prioritise prospecting. Growth through prospecting drives higher incremental value; retargeting often involves users who would purchase anyway.
  • Monitor the learning period. New attribution models require time to optimise. Watch closely for the first 2–4 weeks.
  • Use "Compare Attribution Settings" to see how conversions are reported across different windows simultaneously — invaluable for the transition period.

The Bottom Line

Meta's Incremental Attribution represents a genuine evolution in how paid social performance is measured. If you've ever wondered whether your retargeting campaigns are actually driving sales or just taking credit for them — this answers that question.

The short version: prospecting campaigns will look better under incremental attribution. Retargeting campaigns will look worse. Both of those things are accurate. Adjust your strategy accordingly.