Attribution, Reporting and Testing

On This Page

1. Attribution Settings
2. Apple Privacy Changes
3. Reporting Columns
4. Breakdowns
5. A/B Tests and Lift Studies
6. In Practice
7. Key Terms
8. Key Facts
9. Sources

Attribution Settings

An attribution setting tells Meta how long after someone clicks or views an ad a conversion still counts. The common setting is 7-day click and 1-day view: a purchase within seven days of a click, or within one day of seeing the ad without clicking, is credited to the ad. Video campaigns can also count 1-day engaged view, which credits people who watched a meaningful part of a video.

On January 12, 2026, Meta stopped returning data for the 7-day view and 28-day view windows in its reporting API, following an announcement in October 2025. Accounts that had relied on those longer view windows saw lower reported conversions afterward with no change in actual sales. Compare periods before and after that date with care.

The Compare attribution settings option in Ads Manager shows results side by side under different windows, which helps separate conversions driven by clicks from those credited on views alone.

Apple Privacy Changes

Since Apple introduced App Tracking Transparency in 2021, iPhone apps must ask permission before tracking people across other companies' apps and websites. Many people decline, so Meta sees fewer website actions directly and fills some gaps with modeling. Server-side tracking through the Conversions API, covered in the previous guide, recovers part of that signal.

Reporting Columns

The default columns in Ads Manager rarely match what a business needs. Build a saved column preset with the numbers that answer your questions, for example:

• Amount spent, results and cost per result
• Reach, frequency and CPM
• Link clicks, CTR (link) and landing page views
• Leads or purchases and their value

Landing page views alongside link clicks show how many clicks actually loaded your page. A large gap usually points to a slow site.

Breakdowns

The Breakdown menu splits results by age, gender, placement, region, device and time of day. Breakdowns are useful for spotting patterns and misleading when used to switch segments off one by one. Meta describes a breakdown effect: the system moves spend toward segments that are cheap now to keep the overall cost low, so a segment with a higher average cost may still be lowering the campaign's total cost. Make changes at the ad set level only after testing them.

A/B Tests and Lift Studies

An A/B test in Ads Manager splits the audience so each person sees only one version, then reports which version won and how confident Meta is in the result. Test one variable at a time, such as two offers or two creative concepts.

A conversion lift study withholds ads from a random holdout group and compares conversions between people who could see the ads and people who could not. It answers the question every owner eventually asks: how many of these customers came because of the ads. Lift studies need enough conversions to be reliable, and availability depends on the account.

In Practice

Consider an online furniture retailer whose reported purchases dropped by a third in January 2026 while revenue in its store system held steady. Switching its saved report to the same 7-day click and 1-day view setting for both periods showed click-driven results had barely changed, so the team kept the budget in place.

Key Terms

Attribution setting

The click and view windows Meta uses to credit conversions.

View-through conversion

A conversion credited to an ad that was seen but not clicked.

Breakdown

A report split by a dimension such as age or placement.

Breakdown effect

The tendency for segment averages to mislead because delivery shifts spend toward the cheapest results available now.

Conversion lift

An experiment that measures additional conversions caused by ads using a holdout group.

Key Facts

• From January 12, 2026, the remaining view-through window in Meta reporting is 1-day view.
• Research on large Facebook ad experiments found common observational methods frequently over or understate ad effects compared with randomized tests.

Sources

1. Ads Insights API metric availability updates, Meta for Developers developers.facebook.com
2. User privacy and data use, Apple Developer developer.apple.com
3. Gordon, Zettelmeyer, Bhargava and Chapsky, "A Comparison of Approaches to Advertising Measurement: Evidence from Big Field Experiments at Facebook," Marketing Science, 2019, 10.1287/mksc.2018.1135 doi.org
4. Meta Business Help Center www.facebook.com

Last reviewed October 6, 2026. Platforms change often, so confirm details in the sources above before acting on them. This guide is general information and is not marketing advice for your business.