Meta Pixel, Conversions API and attribution

How do I change the attribution setting for a Meta ad account?

The attribution setting sits on each ad set, not the ad account. Here is how to change it, what each window counts, and what follows delivery.

Published

The attribution setting is not a control of the ad account. It sits on each ad set, in the Conversion section, and it decides which conversions that ad set is credited with. To change it you edit ad sets one at a time, and the closest thing to an account wide change is the comparison view, which reports other windows without touching delivery.

Where the setting lives and what it counts

Meta’s attribution settings, which many still call the Facebook attribution window, are chosen per ad set. For website leads the standard model lets you combine three kinds of credit, and the default is 7-day click with 1-day view and 1-day engagement.

SettingWhat counts as a lead for the adEffect on reported cost per lead
1-day click-througha lead within a day of a link clickfewest leads, highest cost per lead
7-day click-througha lead within a week of a link clickmore leads, lower cost per lead
1-day engage-througha lead within a day of a non-link click, such as a like or a save, or of a short video playadds leads with no link click
1-day view-througha lead within a day of an impressionadds leads credited to a view alone
The default, all three combinedeverything abovethe highest count, so the lowest cost per lead

The full list of windows and what each one includes is in Meta attribution settings, and the view side has its own entry in view-through attribution.

Steps to change the setting on an ad set

Meta’s own setup steps run through the ad set screen:

  1. In Ads Manager click Create and choose Leads, Sales or Engagement as the objective.
  2. At ad set level, in the Conversion section, set the conversion location to Website.
  3. Click Show more options.
  4. Hover over Standard attribution selections and click Edit.
  5. Set click-through to 1 day or 7 days, engage-through to 1 day, and view-through to None or 1 day.
  6. Finish the ad and publish.

The screen can look different if your account has not received the updated engage-through option yet, so expect small naming differences while Meta rolls the change out.

The ad set is the wrong level to be careless at. Splitting your account so that each offer, audience and group of creatives has its own ad set is what makes an attribution choice readable later, which is the case for a small lead gen account structure over a deep one.

What the setting does to delivery

Meta’s help article is explicit on two points. The attribution model you choose informs ad delivery, and standard attribution optimizes delivery for the windows you select. So tightening the window to make Ads Manager agree with your CRM also changes what the algorithm treats as a successful ad set.

That is a real trade rather than a reporting preference. A tighter window tells Meta to look for people who convert quickly after a click, which can suit a short sales cycle and can starve one that closes after a call. Before you edit, ask which of the two you want: an account that optimizes for a different definition of a lead, or a report that shows the same delivery through another lens. The second one is free and reversible.

What you can no longer select

Long windows are gone as ad set settings. Click-through tops out at 7 days, views and engagement at 1 day. Two leftovers remain and are easy to confuse with selectable settings:

  • 28-day click exists only as a column in Compare attribution settings, for reporting.
  • The 7-day and 28-day view windows were removed from the Ads Insights API on January 12, 2026, so any dashboard built on them needs a check.

Meta also offers incremental attribution on the same screen. It optimizes and reports on conversions its models predict the ad caused, which is a different count from the standard one. Meta warns against comparing ad sets that run different models in the same campaign table, and the warning earns its place: a mixed table looks like a performance difference when it is a definition difference.

Comparing windows without changing delivery

  1. Open the Columns: Performance dropdown and keep Performance selected.
  2. Scroll down and choose Compare attribution settings.
  3. Tick the windows you want: 1-day view, 1-day click, 7-day click, 28-day click. Engage-through appears as 1-day Engagement.
  4. Choose All conversions or First conversion.
  5. Click Apply, then export the table if you report on it.

First conversion is the closer match to a CRM that merges repeat submissions from the same person, and a campaign that has not been live for a month only has partial data in the 28-day click column. Both details change the number you quote, so label the column in the client report.

Reading a change in your cost per lead

Cost per lead is spend divided by credited leads. Your budget and the auction set the spend; the attribution setting sets the denominator. Widen the window and the same spend shows more leads and a lower cost per lead. Narrow it and the opposite happens, without a single euro of extra performance.

Practical rules that keep this from turning into a reporting argument:

  1. Fix one setting per client account and write it down. Change it only when the sales cycle really changed.
  2. Compare a period against a period that used the same setting, or re-read the old numbers in the new window.
  3. Keep a source of truth outside Meta. A lead carrying the click identifier that produced it, plus the ad it came from, lets you rebuild the count yourself.
  4. Mark the March 2026 change in older reports. Click-through now counts link clicks only, so engagement moved into engage-through and click-through totals got smaller.
  5. When the CRM and Meta disagree, look at the leads with no click attached before blaming either side. A view-through lead never clicked, so your landing page cannot see it.

That last group is where the practical argument for first-party tracking starts. DropCPL writes a first-party cookie on your own landing page and keeps the click identifier with each lead, so the ad, the creative and the landing page stay attached to the record; the attribution window then becomes a setting you can read against a count you own.

A short checklist before you edit

  • Is the setting going on the ad set you mean to change, rather than on a duplicate?
  • Do you want different optimization, or only a different report?
  • Will you compare like with like, on the same window, at both ends?
  • Does the client report say which window and which date the definition changed?
  • Do your own lead records carry the click identifier and the ad it came from?

Questions on this topic

All questions on this topic →

FAQ

Can I set one attribution window for the whole ad account?+

No. The attribution setting belongs to the ad set, inside its Conversion section, so an account wide change means editing each ad set or using the campaign duplication flow. For reporting across the account you can add the compare columns, which do not touch delivery.

Does changing the attribution setting change how the campaign spends?+

Yes, when you change the setting the ad set optimizes for. Meta says the model you pick informs delivery as well as how conversions are credited, and standard attribution optimizes for the windows you select. If you only want a different view of the same delivery, use the comparison columns instead.

Why did my reported cost per lead change without any budget change?+

Cost per lead is spend divided by credited leads, and the attribution setting sets the denominator. Wider windows credit more conversions to the same spend, so the reported cost per lead falls. Nothing about the auction changed.

Is the 28-day click window still available to select?+

Not as an ad set setting. The maximum for click through is now 7 days, and 28-day click survives only as a reporting column in Compare attribution settings. The 7-day and 28-day view windows were removed from the Ads Insights API on January 12, 2026.

Ready to find out what
converts, for real?

Free to start.

Start free →