Facebook Ads Attribution Windows: How to Track What Actually Converts in 2026

Two ad accounts running the same offer, same creative, same budget. One reports 40 conversions at $18 CPA. The other reports 22 conversions at $33 CPA. Same reality, different attribution settings. The first account counts view-through conversions and uses a 7-day click window. The second uses 1-day click only. Neither number is wrong. Both are incomplete.

Attribution windows decide which conversions show up in your reports. They shape the data the algorithm uses to optimize delivery. They determine whether your CPA looks profitable or bleeding. And most media buyers never change them from the default, never cross-check them against backend data, and never realize their "winning" campaigns might be taking credit for conversions that would have happened anyway.

What attribution windows are and what they control

An attribution window is a countdown timer. It starts when someone interacts with your ad, whether a click or a view, and runs for a set number of days. Any conversion event that fires inside that window gets credited to the ad. Anything outside the window disappears from your Meta reports.

Meta offers four attribution settings at the ad set level:

• 7-day click, 1-day view (the default). Conversions from clicks within 7 days plus conversions from ad views within 1 day.
• 7-day click. Click-only conversions within 7 days. No view-through credit.
• 1-day click, the strictest setting. Only same-day click conversions. iOS 14 opted-out users get this regardless of your setting.
• 1-day click, 1-day view. Same-day clicks plus same-day views. A middle ground that still includes view-through.

The setting lives inside each ad set under "Attribution setting" in the Optimization and delivery section. Changing it does not rewrite historical data. It affects reporting and optimization from that point forward only.

Click-through vs view-through: what each means

Click-through attribution is straightforward. Someone clicked your ad, then converted. There is a direct causal chain: they saw an ad, they acted on it, they bought something. Even with a multi-day gap between click and purchase, the intent was expressed through a deliberate action.

View-through attribution is less clear. Your ad appeared in someone's feed. They scrolled past without clicking. Later that day, they visited your site and bought. Meta credits your ad with that conversion. The ad may have planted a seed. Or they were going to buy regardless, and Meta showed them the ad because it already flagged them as a likely buyer.

View-through inflates conversion counts. When you compare Meta reported conversions against backend sales data, view-through conversions account for 15 to 40 percent of reported totals. Strip them out and CPA rises, ROAS drops, and the performance picture changes.

View-through still has value for brand awareness, video campaigns, and top-of-funnel work. It captures the indirect influence ads have on purchase behavior. You go wrong when you treat view-through numbers and click-through numbers as equal inputs to profitability math.

The right setting depends on your conversion cycle

The attribution window should match how long your customers take to buy. Use a window that is too short and you lose data. Too long and you take credit for conversions you did not cause.

Use 7-day click for almost everything. It feeds the algorithm enough conversion data to optimize. Two exceptions: app installs where the action is instant, and B2B where the real conversion (a signed deal) happens weeks or months after the click.

iOS 14 and the attribution collapse

Apple's App Tracking Transparency framework broke Facebook attribution for good. Roughly 75 to 85 percent of iOS users opt out of tracking. For those users, Meta cannot follow behavior across apps and websites.

Four consequences hit your data:

• Forced 1-day click window. Opted-out users can only be attributed within 1 day of a click, regardless of your ad set setting. If an opted-out iOS user clicks your ad and buys on day 3, that sale vanishes from Meta reports.
• No view-through for opted-out users. View-through attribution requires cross-app tracking, which ATT blocks. View-through numbers in your reports come only from Android users and the 15 to 25 percent of iOS users who opted in.
• Reporting delays. Conversions from opted-out users rely on Apple's SKAdNetwork or Meta's Aggregated Event Measurement, both of which batch data with delays of up to 72 hours.
• Event prioritization. You are limited to 8 conversion events per domain per pixel. Meta uses these ranked events for aggregated reporting. If Purchase is your top event, lower-priority events like Add to Cart may under-report.

Meta under-reports conversions by 20 to 50 percent in iOS-heavy audiences. If 60% or more of your traffic comes from iOS (common in the US, UK, Australia, and Western Europe), your reported conversions miss a large chunk of actual sales.

How to close the data gap

You need multiple tools working together to close the gap:

1. Conversions API (CAPI). Server-side tracking sends conversion events directly from your server to Meta, bypassing browser restrictions. CAPI recovers 20 to 35 percent of the conversions that browser-only pixel tracking misses. If you run CAPI alongside the pixel, Meta deduplicates events and uses the more complete dataset. In 2026, running ads without CAPI means the algorithm has less data to optimize with. Your CPA pays the price.

2. UTM parameters and GA4. Tag every ad URL with UTM source, medium, campaign, and content parameters. GA4 uses last-click attribution by default, which gives you a more conservative view of which campaigns drive conversions. Compare Meta reports against GA4 weekly. Your real numbers sit somewhere between the two.

3. Backend revenue matching. Export your Meta campaign data and match it against actual backend sales by date. If Meta says you got 50 purchases yesterday and your Shopify dashboard shows 62, you know Meta is under-reporting by 19 percent. Track this ratio over time. It becomes your correction factor for budget decisions.

4. Post-purchase surveys. Add a "How did you hear about us?" field to your checkout or thank-you page. People forget and people lie, but surveys catch channels that attribution models miss: word-of-mouth referrals from someone who saw your ad, podcast mentions, offline conversations.

View-through conversions: include or exclude?

It depends on what you use the data for.

Include view-through when you measure brand lift, run awareness campaigns, or report on top-of-funnel activity. Video campaigns generate large view counts that plant purchase intent. Stripping view-through from a video awareness campaign makes it look dead when it might be working.

Exclude view-through when you make budget decisions based on CPA or ROAS. If you see "$25 CPA" and 30 percent of those conversions are view-through, your real CPA is closer to $36. That changes the math on whether to scale.

A practical approach: report with view-through included, but make scaling decisions on click-through only. Pull both columns into your reporting dashboard. Over time you build an intuition for how much view-through to trust in your specific account.

The comparison columns most buyers overlook

Ads Manager lets you compare results across attribution settings without changing your actual setting. Go to Columns → Customize Columns → Compare Attribution Settings. This adds side-by-side columns showing conversions under different windows for the same data.

You read the delta between columns and see how your attribution setting shapes perceived performance:

• 7-day click shows 50 conversions, 1-day click shows 38. Those 12 missing conversions (24%) happened between day 2 and day 7. Real people clicked and bought later. A 1-day window hides them.
• 7-day click + 1-day view shows 65, 7-day click alone shows 50. Those 15 extra conversions (23%) are view-throughs. Cross-check a random sample against your backend to see how many are genuine.

Run this comparison once a month. The ratios shift as your audience mix, creative approach, and iOS/Android split change over time.

Common attribution mistakes

Mixing attribution windows across ad sets. If one ad set uses 7-day click and another uses 1-day click, you cannot compare their CPA. The first will always look better because it counts more conversions over a longer window. Pick one setting and apply it to every ad set in the account.

Comparing Meta numbers against Google Ads without normalizing. Google Ads defaults to 30-day click attribution. Meta defaults to 7-day click plus 1-day view. Comparing CPA between the two platforms without aligning the attribution window produces nonsense.

Ignoring modeled conversions. Meta marks some conversions as "modeled," meaning statistically estimated rather than tracked. These fill the iOS data gap. Some advertisers strip modeled conversions from reports to be "conservative." That under-counts by 15 to 30 percent and starves the algorithm of signal. Keep modeled conversions in. The algorithm uses them for optimization whether you report on them or not.

Resetting attribution mid-campaign. Changing the attribution window on a running ad set does not rewrite historical data. But it changes what the algorithm optimizes for going forward, which can trigger a learning phase reset. Change attribution settings when you duplicate ad sets, not on live ones.

Trusting one source. Meta reports, GA4, backend data, and post-purchase surveys tell different stories. None of them is complete. Triangulate: if three out of four sources agree that Campaign A outperforms Campaign B, act on it. If they disagree, dig deeper before spending more.

Setting up attribution correctly: step by step

1. Audit your current settings. Open Ads Manager, select all active ad sets, click Edit. Check the Attribution Setting field. If different ad sets have different windows, you have inconsistent data. Note what each one uses.
2. Pick one window for the account. 7-day click for most advertisers. Exclude view-through unless you specifically need it for awareness measurement.
3. Standardize going forward. Create a naming convention or checklist that includes attribution setting verification before any ad set goes live.
4. Turn on Compare Attribution Settings. Add the comparison columns to your default reporting view. Review the delta between windows monthly.
5. Implement CAPI. If you have not done this already, it is the single highest-impact change you can make for data quality. Meta estimates a 13 to 20 percent improvement in attributed conversions after CAPI implementation.
6. Build a cross-platform dashboard. Pull Meta data, GA4 data, and backend revenue into one view. Calculate your Meta under-reporting ratio weekly. Use it as a correction factor for budget decisions.

What to remember

• Attribution windows determine which conversions show up in your reports. They directly affect your CPA, ROAS, and the data the algorithm uses to optimize.
• 7-day click is the right setting for most advertisers. It captures the full post-click conversion cycle without padding numbers with view-through.
• View-through conversions inflate reported performance by 15 to 40 percent. Include them in reports for context but make budget decisions on click-through data.
• iOS 14 ATT forces 1-day click attribution for opted-out users regardless of your setting. Expect 20 to 50 percent under-reporting in iOS-heavy audiences.
• Conversions API recovers 20 to 35 percent of lost conversion data. Running without it in 2026 means worse optimization and worse CPA.
• Never compare campaign performance across different attribution windows. Standardize to one setting across all ad sets.
• Triangulate: Meta reports plus GA4 plus backend revenue plus post-purchase surveys. No single source is complete.