How Meta attribution settings change your results

How do Meta attribution settings affect your results?

Meta's attribution setting decides after which interaction, and within how long, a purchase on your site gets linked to your ad. The wider the window, the more sales and the higher ROAS Ads Manager reports; the narrower it is, the lower the numbers. The setting also affects which behaviour delivery optimizes for, not just the report. So the same campaign both looks different and runs differently under different attribution settings.

That is why the answer to "why did our ROAS drop?" sometimes lies not in the campaign but in a changed attribution setting, or in comparing ad sets that use different settings side by side.

Which attribution models does Meta offer?

Meta lets you choose an attribution model at ad set level. According to Meta, these models both inform ad delivery and determine how conversions are credited to your ads:

• Standard attribution: optimizes delivery for selected time windows and user behaviours. You can choose whether to credit conversions based on impressions, clicks and/or video plays. Attribution settings are chosen only in this model.
• Incremental attribution: uses models that predict whether a conversion was caused by an ad, and optimizes and reports on incremental conversions.
• Custom attribution: may not be available to everyone yet. It lets you share attribution data from your external analytics tool with Meta via the Conversions API or a supported analytics partner, so delivery optimizes towards your own attribution logic.

What are the standard attribution settings?

The standard settings Meta supports for website and in-store conversions are:

Window	What it counts	Options
Click-through	Events after a link click on your ad	1 day or 7 days
View-through	Events after an impression of your ad	1 day
Engage-through	Events after a non-link click action on your ad	1 day

Engage-through covers any clicks on your ad except link clicks. For video ads, a 5-second play also counts (or 97% of the video if it is shorter than 5 seconds). Meta notes that some accounts may still use prior versions of click-through and engage-through attribution while this rolls out.

A combination you will often see in ecommerce accounts is "7-day click and 1-day view". Under that setting, someone who clicks your ad and buys six days later is attributed to the ad, and so is someone who sees the ad, does not click, and searches for your brand to buy the next day.

How does the attribution window change reported sales?

Widening the window does not increase actual sales; it only increases the sales attributed to your ads. The table below shows the difference with a hypothetical calculation.

Example: one campaign, three windows

Example: a coffee brand spends USD 1,000 in one week. Reading the same period under different attribution settings changes the picture like this (figures are hypothetical):

Attribution setting	Attributed purchases	Attributed revenue	ROAS
1-day click	60	USD 2,400	2.4
7-day click	85	USD 3,400	3.4
7-day click + 1-day view	110	USD 4,400	4.4

All three rows describe the same spend and the same real sales. The difference is which sales count as "belonging" to the ad. View-through sales include people who saw the ad but did not click; some of them may be existing fans who would have bought anyway. So be careful when you compare a wide-window ROAS with your profit target. We explain how ROAS is calculated in our ROAS guide.

Does the attribution setting affect delivery too?

Yes. Meta states that the attribution model, and the windows chosen in the standard model, inform delivery. In practice: if you choose only a click window, the system focuses on people likely to click and then buy. If you add a view window, buying after seeing the ad without clicking also counts as success, and the system looks for those people as well.

So do not change your attribution setting often as if it were a report filter. Set it once according to the campaign's business goal and stay consistent. Meta also stresses that results cannot be compared in the campaign overview table across ad sets with different attribution models, because each model uses different counting mechanisms. If you need to compare, use the compare attribution settings feature in Ads Manager.

Which attribution setting should you choose?

The right setting depends on the question you want answered:

1. If you want to see sales the ad drove directly, focus on the click window. Basing budget decisions on click-through results is more conservative and usually safer.
2. If your products take several days to decide on (furniture, electronics), 7-day click gives a more realistic picture than 1-day click.
3. For video-heavy, discovery-led campaigns, view-through and engage-through windows make the people who buy without clicking visible, but read those results separately.
4. If you care whether the ad truly creates extra sales, consider incremental attribution. According to Meta, it helps focus campaigns on outcomes more likely to be directly driven by an ad.
5. If you run your own attribution system or a supported analytics partner, custom attribution may be an option. According to Meta, it can only be enabled on new campaigns and usually needs 1-2 weeks of high-quality signals before you can create campaigns with it.

How do you standardise the attribution setting across a team?

If the agency, the in-house team and management discuss the same ROAS figure under different settings, meetings stop being useful. Set a simple rule: every sales ad set in the account uses the same attribution setting, and that setting is written at the top of every report. Set your target ROAS under the same setting too; do not compare a result measured on 7-day click and 1-day view with a target calculated on clicks only. If you need to change the setting, log the date so you can interpret the break in the numbers later.

Why does Meta reporting differ from your analytics tool?

Attribution is one of the biggest reasons Meta's numbers differ from tools such as Google Analytics. According to Meta, the main differences are: Meta can count view-through conversions while many third-party tools cannot; Meta's people-based measurement can track cross-device conversions; ad blockers can stop the Pixel firing; and tools relying on referrer URLs can miss a conversion when someone buys in a new tab.

Events Manager and Ads Manager also show different numbers: Events Manager shows most events received, whether or not they relate to an ad, while Ads Manager only shows deduplicated events attributed to people who were shown your ad. To strengthen your measurement setup, see our GA4 ecommerce tracking guide.

Key takeaways

• The attribution setting decides which interaction and what time window can link a purchase to your ad.
• In the standard model you can choose 1 or 7 days for clicks and 1 day for views and engagement.
• Wider windows report higher ROAS but do not create extra sales; read them carefully for profit decisions.
• The attribution model also shapes delivery, so do not change it often and do not compare different models in one table.
• If you want to know the ad's true impact, consider incremental attribution.

If you want to find out how much of your reported ROAS really comes from your ads, you can ask the Performetic team for a free growth analysis.