Meta Attribution Changes 2026: What to Tell Clients

Meta made two measurement changes in 2026 that can make reported conversions fall without any change in real performance. On January 12, 2026 it removed the 7-day view and 28-day view attribution windows. On March 3, 2026 it announced that click-through attribution now counts only link clicks, with likes, shares, saves and other interactions moved into a new "engage-through" category. Billing did not change, but any report that compares periods across these dates is no longer comparing like with like.

This guide explains what changed, how to check the effect in your own accounts, and gives you wording to send to clients before they ask.

Timeline

What the default looks like now

After the March change, the default attribution setting is 7-day click, 1-day engage-through and 1-day view. The longest view window available is one day. Engage-through applies by default to campaigns optimizing for website or in-store conversions, and advertisers can turn it off by choosing "None" in the attribution settings.

Coverage of the March change describes it as bringing Ads Manager closer to third-party tools such as Google Analytics, which credit conversions to link clicks.

Why your numbers may have dropped

• Fewer conversions credited to views. The January removal took away credit for conversions that happened days after someone only saw an ad. Long-consideration purchases, remarketing and awareness or video campaigns were the most exposed, according to vendors who analyzed the change.
• Some click credit moved to engage-through. A conversion that used to count as click-through because someone liked or saved the ad now sits under engage-through. If that person converted on day two to seven, it falls outside engage-through's one-day window and disappears from the attributed total.
• The size varies. Third-party estimates often quote drops of 15 to 40 percent for affected accounts, but the real figure differs by account, objective and audience. Never promise a number to a client. Measure your own.

What it means for your reports

1. Comparisons across the change are mixed. A month-over-month "decline" in early 2026 can be a measurement effect rather than a performance one.
2. Year-over-year comparisons are affected too. 2026 conversions are counted differently from 2025 conversions.
3. Old targets may be stale. A ROAS or cost per result target set under the old rules may need re-basing.
4. Tools cannot see the definition change. Any tool that shows a percentage change between periods, MetriQuill included, calculates it from the numbers in the export. It cannot know that the counting rules moved in between, so annotate the report yourself.

How to handle it, step by step

1. Find your break points. Export daily data and look for the dates where conversions stepped down. They may differ by account.
2. Show both columns. Add engage-through next to click-through in Ads Manager so you can see the whole picture. Treat them as two lenses. Do not simply add them together to "restore" the old number, because summing overlapping categories can double count.
3. Set a new baseline. Compare post-change periods with other post-change periods, such as April with May, and check Meta's numbers against independent data: store orders, CRM leads or Google Analytics.
4. Re-base targets. Recalculate expected cost per result from post-change data and agree the new targets with the client.
5. Annotate the report. Mark the change dates on charts and tables, and state the attribution setting once in the appendix.
6. Tell the client first. Use the wording below.

Wording you can send to clients

Meta changed how it counts conversions twice this year. In January it removed the longer view-through windows, and in March it narrowed "click-through" conversions to actual link clicks and moved other interactions into a separate category. Meta has said billing is unaffected, and your spend is exactly what we planned. The effect is on reporting: some conversions that used to be credited to ads are not credited now, so results in Meta can look lower without any drop in real sales or leads. To keep the picture honest, we now compare Meta's numbers with [your store orders or your CRM], and we measure progress against a new baseline starting in [month]. The strategy has not changed. The yardstick has.

And a one-line footnote for the appendix:

Meta's conversion attribution changed in January and March 2026. Figures before and after those dates are not directly comparable. Attribution setting used: [setting].

What not to do

• Do not re-base numbers quietly. If the client finds out later, every earlier number becomes suspect.
• Do not blame Meta for real problems. Check independent data first. A genuine drop in store sales or qualified leads is not a measurement effect.
• Do not promise the numbers will come back. The removed windows are gone.
• Do not change the attribution setting mid-report to make a comparison look better. Pick one setting per client and keep it stable so periods stay comparable.

Frequently asked questions

Did Meta change what I pay? Meta has said billing is unaffected by the March change. The change is in how conversions are classified and reported.

Is 7-day click still available? Yes. It remains part of the default setting.

Should I turn engage-through off? There is no universal answer. Some advertisers keep it for prospecting and switch it off for remarketing to stay conservative. Choose one approach per client and stay consistent.

Do CSV exports reflect the change? Exports show conversions under the attribution setting used for the report, so files from before and after the changes were counted under different rules. See how to export Meta Ads data to CSV.

How does this affect benchmarks? Published benchmarks that mix pre-change and post-change data are not perfectly comparable with your current numbers. Use them as guardrails. More in Facebook Ads benchmarks 2026.

Sources

• Meta for Business: Simplifying Ad Measurement for a Social-First World (March 2026)
• Search Engine Land: Meta introduces click and engage-through attribution updates
• PPC Land: Meta rewrites click attribution rules
• Jon Loomer Digital: Click-through attribution now requires a link click
• Dataslayer: Meta Ads attribution window removed, January 2026
• Conversios: Meta attribution window changes 2026

The January change is documented mainly through developer announcements and analytics vendors, so confirm the details in your own account.

Related reading

• Meta Ads Reporting for Agencies: 2026 Guide
• How to present Facebook Ads results to clients
• Meta Ads "Results" and cost per result explained
• Facebook Ads benchmarks 2026: CTR, CPC and CPL

Last reviewed September 2026. Meta may adjust these rules again, so check Ads Manager and Meta's announcements for the latest.