Skip to main content

Attribution Window: What It Is and How to Set It per Traffic Source

Our write-up of a story first reported by Voluum Blog

Attribution Window: What It Is and How to Set It per Traffic Source

Meta reports 42 sales. A tracker reports 31. The affiliate network pays for 33. Nobody is lying and nothing is broken: each system runs its own rule about how long after a click a sale still counts as the ad's.

That rule is the attribution window. It decides which campaigns look profitable, which get paused, and how much credit each traffic source takes for the same conversion.

How the window works

The window is a time limit. It opens when someone interacts with an ad and closes a set number of days later. A conversion inside the window is credited to the ad; one outside it still happened, but the ad gets nothing for it. A click on a Meta ad on Monday and a purchase nine days later never appears in Ads Manager as a result of that ad under the default 7-day click window, while a tracker with a longer window still matches it to the original click.

Platforms run several windows side by side. The click-through window starts at the click and is the one that matters for direct response. The view-through window starts when someone sees the ad without interacting. The engage-through window sits between them, covering video views and social actions such as likes and shares. A click is the clearer signal of intent, so for affiliate offers, lead gen and most performance campaigns, judge on click-through conversions first and treat view-through as a bonus to verify. Meta's Compare Attribution Settings shows the two side by side; where view-through makes up a large share of reported results, reported ROAS is likely more optimistic than the bank account.

Defaults by traffic source

Meta runs 7-day click, 1-day engage-through and 1-day view under standard attribution, set per ad set. The click window can be 1 or 7 days, view-through 1 day or off. Since March 2026 click-through attribution requires a link click, with likes, shares and other non-link clicks moved to engage-through. From January 12, 2026, Meta stopped returning 7-day and 28-day view windows in reporting, so older comparisons will not line up with current numbers.

Google calls it the conversion window and sets it per conversion action: 30 days for click-through by default, running from 1 to 90 days depending on the conversion source, 1 day for view-through and 3 days for engaged-view conversions on video. The standard Conversions column credits each conversion to the day of the ad interaction, not the day of the sale, so last week's figures can keep growing.

TikTok defaults web campaigns to 7-day click and 1-day view at ad group level, with click windows of 1, 7, 14 or 28 days and view-through off, 1 or 7 days. App campaigns add a separate engaged-view window. Taboola counts conversions within 30 days of a click and within 24 hours of a viewable impression by default, both changeable per conversion rule; when a user both clicked and viewed, the click wins. Other native networks follow similar logic with defaults that rarely match Meta or Google.

Many push, pop and native networks run no pixel-based window at all. They receive conversions through a server-to-server postback carrying the click ID from the original visit, so the effective window depends on how long the network keeps its click data and on the tracker.

What the window does to reported ROI

One campaign through four windows. Spend of $1,000 on an offer paying $40 per conversion, with 50 conversions over 30 days: 20 within a day of the click, 15 between day 2 and day 7, 10 between day 8 and day 30, and 5 who only saw the ad and converted the same day. On a 1-day window the campaign looks like it is losing money and gets paused. On a 30-day window it looks worth scaling. Neither number is wrong. Comparing a 7-day figure from one source with a 30-day figure from another is where it breaks.

There is no universally correct window, only one that fits the offer. Measure the real conversion lag in the tracker: if 90% of conversions arrive within three days, a 30-day window mostly adds noise, and if a large share arrives after a week, a 7-day window hides results. Sweepstakes, app installs and cheap e-commerce convert fast; lead gen, high-ticket products and iGaming deposits often take longer, particularly when the first deposit comes days after registration. Compare before changing, and leave live campaigns alone, since on Meta the attribution setting also guides delivery and switching mid-flight can reset learning. Put sources on the same window before judging one against another.

Even with matching windows, platform and tracker figures rarely agree to the last conversion. Platforms count view-through and engage-through conversions from people who never clicked, estimate some conversions when privacy settings or ad blockers cut the signal, and close their windows earlier than a tracker that keeps matching late conversions to the original click. Sending conversions back through a Conversion API does not override the platform's window: a sale reported on day 10 against a 7-day click ad set arrives but is not credited under that setting.

Voluum matches each conversion to the click that caused it through the click ID, whether it arrives by S2S postback, conversion pixel or manual upload, and keeps attributing for up to 180 days. Every traffic source runs through the same rules, giving one window for comparing Meta, Google, TikTok, native and push side by side, and Automizer can send recorded conversions back to the ad platforms. Late conversions can be reported by visit timestamp, the default, or by postback timestamp, which lines up more closely with what an affiliate network or operator reports for a given day; the option sits in General Settings and applies per user.

TG X Copied

Project sponsors

Companies that keep this analytics open