Meta's attribution system doesn't just measure performance. It manufactures it. Pick the wrong window and your ROAS looks great while your revenue stays flat. Every media buyer has seen it — the account where pausing Meta ads seemed to "hurt" sales that were already happening anyway.
In this post:
- How Meta's three attribution windows work and what each actually measures
- Why view-through attribution is the most common source of inflated ROAS
- Which window to use for prospecting vs. retargeting campaigns
- What changed in 2026 — and why reported conversions dropped 15–40% overnight across thousands of accounts
- How to get cleaner attribution signals before misallocation compounds
What Meta's Attribution Windows Actually Measure
Attribution is an accounting question: which ad gets credit for a conversion? Meta's answer depends on the window you set.
Three windows exist in 2026:
| Window | What it credits |
|---|---|
| 7-day click | User clicks your ad, buys within 7 days |
| 1-day click | User clicks your ad, buys within 24 hours |
| 1-day view | User sees your ad without clicking, buys within 24 hours |
The default in Meta Ads Manager is 7-day click + 1-day view for most conversion objectives. That sounds reasonable. The problem is the "1-day view" half.
View-through attribution credits a conversion to your ad even when the user never clicked — they simply saw it. That's a legitimate signal for brand awareness campaigns. For bottom-of-funnel conversion campaigns, it routinely takes credit for purchases that would have happened without the ad. According to Meta's own attribution documentation, view-through attribution is best suited for awareness and reach objectives — not purchase campaigns where click intent matters.
The View-Through Trap
Here's how inflated ROAS compounds quietly.
You run a retargeting campaign against your existing customer list. A customer who bought last month sees one of your ads in their feed. They don't click. Three hours later, they come back to your site through a Google search and buy again. Meta claims that conversion.
Across high-retention DTC brands, audits have found that 1-day view attribution quietly consumed 18–34% of reported conversions — conversions belonging to repeat customers who were already in the purchase loop regardless of ad exposure. Research suggests that attribution window problems can inflate or deflate conversion counts by 30–60%, which directly distorts how budgets get allocated and which campaigns look healthy.
This isn't a new problem. But 2026 made it impossible to ignore.
What Changed in 2026
Two structural changes hit Meta attribution this year, both catching teams off-guard.
January 2026: Meta permanently removed the 7-day view and 28-day view windows from the Ads Insights API. Accounts reporting on those windows saw conversions drop 15–40% overnight — not because campaigns degraded, but because the measurement methodology changed beneath them. The revenue was still there. The reported numbers weren't.
March 2026: Meta redefined what counts as a "click" for click-through attribution. Previously, any interaction with an ad — likes, shares, saves, reactions — counted. Now, only outbound link clicks that actually send users to your website count. Social engagement moved to a separate engage-through category with a 1-day window only.
The practical effect: if your performance appeared flat or declining in Q1 2026, attribution methodology is the first place to audit, not creative or spend.
Prospecting vs. Retargeting: Use Different Windows
The most important attribution decision isn't which single window to use globally — it's that prospecting and retargeting campaigns need different measurement logic.
Prospecting (cold audiences): Use 7-day click. Cold audiences have longer consideration cycles. A prospect who sees your ad Monday might click, compare options, and buy Friday. A 1-day click window makes prospecting look broken when it isn't. Last-click attribution consistently understates prospecting's contribution — but prospecting is where your customer acquisition actually happens. Without a funded prospecting pool, your retargeting audience shrinks to zero.
Retargeting (warm audiences): Use 1-day click, and disable view-through. Warm audiences are already in the funnel. They convert faster. The 7-day click window here inflates ROAS by capturing organic return visits that would have happened regardless. View-through is worse: retargeted customers frequently complete purchases through direct or organic channels anyway, and Meta's attribution claims the credit.
A clean setup separates these signals rather than averaging them together. If your retargeting and prospecting campaigns report into the same attribution setting, you're reading a blended number that's useless for budget decisions.
Clean Signals, Not Big Numbers
The goal of attribution isn't to maximize what Meta shows you. It's to understand which spend is actually driving revenue.
Use the Conversions API. Meta's pixel has gaps — browsers block cookies, iOS restrictions suppress tracking, same-device multi-session journeys get fragmented. The Conversions API sends server-side events directly to Meta, closing the gaps the pixel misses. This is the single highest-leverage tracking improvement available. More on the full setup in Meta Conversions API: The 2026 Setup Guide.
Compare in-platform to independent sources. Meta Ads Manager will always report more conversions than your Shopify dashboard or CRM. That gap is normal — it reflects multi-channel overlap. What matters is the ratio. If Meta claims 3× more conversions than your source of truth, view-through attribution is doing most of the work.
Match the window to your buyer's timeline. A $30 impulse purchase converts in hours. A $2,000 B2B service converts in weeks. Meta's defaults were built for the average, which fits no one perfectly. bulk reads live account attribution data to surface where window mismatches are distorting spend decisions — before the budget has already been misallocated. The connection between accurate attribution and campaign ROI is direct: see Meta Ads Automation ROI: What AI Agents Actually Deliver for how this plays out in practice.
Compare Meta-reported conversions to your actual order count for the same period. If Meta is claiming 30%+ more than your CRM or checkout system, check whether view-through attribution and retargeting audience overlap are the source.
The Number That Actually Matters
Attribution confusion starts as a reporting problem. It becomes a budget problem the moment you let it drive decisions. Over-crediting retargeting starves prospecting. Leaving view-through on inflates ROAS without growing revenue. The January 2026 changes didn't create these problems — they just removed the option to avoid confronting them.
Pick the window that matches how your product is actually bought. Separate prospecting and retargeting measurement. Cross-reference Meta's numbers against your actual backend. The ROAS that matters isn't the one Meta shows — it's the one you can verify.
bulk reads live Meta account attribution data to surface where signal mismatch is costing spend. See how it works →