An attribution window is the fixed period after someone clicks or sees your ad during which a later conversion still gets credited to that ad. It sets the deadline for credit, so it decides which conversions count for a campaign and which fall outside the measured timeframe, unlike the attribution model, which decides how credit is split among touchpoints.
Choose the window and you choose how good every ChatGPT Ads, Google, or Meta campaign looks in your reports. Set it too short and you cut budget from campaigns that were quietly working, because delayed conversions never land inside the window you measured.
Attribution windows set the measurement boundary an advertising platform uses to connect a conversion back to an earlier ad click or impression. Most platforms run two at once: a click-through window that starts on a click, and a view-through window that starts on an impression. In ChatGPT Ads, OpenAI lets advertisers choose a click-through window and adds a fixed one-day view-through window measured from an ad impression.
The window is a per-campaign setting you control in the ads manager. The platform stamps each ad interaction with a time, then keeps a later conversion only if it arrives before the window closes. Click windows run longer than view windows because a click signals stronger intent than a passive impression.
Attribution windows work alongside an attribution model: the window decides which touchpoints are eligible, and the model decides how credit is divided. On a surface like ChatGPT Ads, where many conversions complete off-session and on another device, the window you pick shapes how much delayed activity your reports capture. AirOps helps teams track AI search visibility and the offsite signals behind these journeys.
Resources: a practical guide to running and measuring ChatGPT Ads campaigns
An attribution window runs the same loop every time a conversion fires, matching it against recent ad interactions and checking the clock.
Set the window: You choose the click and view windows per campaign in the ads manager before traffic runs.
Record the touch: When a user clicks or sees your ad, the platform logs a timestamp and an identifier such as a click ID or cookie.
Capture the conversion: Your pixel or server-side API sends the conversion event, like a purchase or lead, back to the platform.
Check the clock: The platform compares the conversion time against the touch time and keeps it only if it falls inside the window.
Assign credit: If both a click and an impression qualify, the click wins the credit, and the conversion appears in your report.
The reported total tells you how many conversions the platform could tie to your ads within the rules you set. It does not tell you about conversions that happened later, on another device, or through a path the pixel never saw.
The attribution window you set changes which campaigns look profitable, and that changes where you move budget next month. Treat it as a reporting lever that quietly rewrites your numbers. Pick the wrong one and you can defund a channel that drives real revenue, or overpay for one that only looked good on paper.
Budget follows reported ROAS: A window that ends too early starves ChatGPT Ads of credit for consideration-stage conversions, so the channel looks weak in the dashboard and loses spend it actually earned.
Cross-channel comparisons break: When one platform reports on a 7-day click window and another on 30 days, the same sales get counted differently, and your side-by-side comparison becomes meaningless.
Forecasts depend on window maturity: Conversions keep arriving for days after the initial click, so judging a campaign before its window matures understates real performance and triggers premature budget cuts.
SEO managers use attribution windows to judge whether organic and AI-search landing pages convert within a realistic consideration period before reallocating effort.
Growth marketers use attribution windows to align ChatGPT Ads reporting with their sales cycle so delayed conversions still credit the campaign that started them.
Demand gen leads use attribution windows to standardize how every paid channel counts a lead before comparing cost per acquisition across the mix.
The click-through window credits conversions after an active click, while the view-through window credits conversions after an impression alone, and because a click shows stronger intent it almost always gets a longer window and priority when both qualify for the same conversion event.
A campaign's conversions accumulate across the full length of its window, so numbers read before the window closes are incomplete and will keep rising as late conversions land, which makes patience essential before any verdict.
An attribution window always counts forward from the ad interaction toward the conversion, so the clock starts at the click or impression and never reaches backward to credit activity that preceded the ad entirely.
Credit delayed conversions that a short window would silently drop.
Compare ChatGPT Ads, Google, and Meta on matched windows for honest cost per acquisition.
Capture view-through impact that click-only reporting would miss entirely.
Protect working campaigns from premature cuts driven by immature window data.
Align reporting with your real sales cycle instead of a platform default.
Lean on OpenAI's 2026 analysis of global campaigns, where 52.7% of eligible one-day view-through conversions occurred within an hour of the matched ad impression.
Match the window to your sales cycle: longer consideration purchases need 14 or 30-day click windows so slow conversions still count.
Keep the view-through window short: a brief view window limits over-crediting impressions that users barely noticed.
Standardize windows across channels: report every paid platform on the same click window before you compare performance.
Wait for window maturity: let the full window pass before judging cost per acquisition, and steer early with clicks and engagement.
Separate view-through from click totals: keep view-through conversions out of your core conversion count so they do not inflate ROAS.
Document your settings: record each campaign's window so later analysis knows which rules produced the numbers.
Avoid changing the attribution window mid-campaign and then comparing before-and-after results as if nothing moved. The window is part of your measurement, so shifting it rewrites history and makes the two periods impossible to read against each other. Lock the window before launch and change it only with a clear reason recorded.
AirOps: tracks your AI search visibility and the offsite signals that shape ChatGPT-influenced conversion journeys, giving context for what your attribution windows can and cannot capture.
OpenAI Ads Manager: sets the per-campaign click and view windows for ChatGPT Ads and reports the conversions that fall inside them.
Google Analytics 4: lets you configure lookback windows and compare how different windows reattribute conversions across all your channels.
Find your current windows: open each ad platform and GA4 and write down the click and view windows every campaign is using right now. Most teams have never checked, and defaults vary by platform.
Map your sales cycle: pull the typical time from first touch to purchase for your product so you know how long a window actually needs to be. A quick cohort pull from past orders is enough to start.
Align the settings: set a consistent click window across channels that matches that cycle, and keep view-through windows short and reported separately.
Install proper measurement: add the conversion pixel and server-side API for ChatGPT Ads so conversions can match to clicks inside the window.
Review after maturity: wait for the longest window to close, then compare cost per acquisition across channels and adjust windows only with evidence. Treat the first mature read as your real baseline before drawing conclusions.
An attribution window is the time limit a platform uses to credit a conversion to an earlier ad click or impression.
Platforms run separate click-through and view-through windows, and the click usually wins when both could claim the same conversion.
A window only counts conversions it can match and that arrive before it closes, so off-session and cross-device sales often go uncredited.
Mismatched windows across channels produce false comparisons that can defund a campaign which was genuinely profitable.
Setting the window to your real sales cycle and reading it only after it matures is where accurate budget decisions come from.
An attribution window and an attribution model answer two different questions, and you need both to read a report correctly. The window is a time limit: it decides which ad interactions are even eligible to receive credit for a conversion, based on how long ago they happened. The model is a credit-sharing rule: once the window has decided which touchpoints qualify, the model decides how much of the conversion each one gets, whether that is all of it to the last click, all to the first, or a split across the path. You can keep the same 30-day window and still see completely different channel numbers depending on whether you run last-click or a data-driven model. In practice you set the window first to define the eligible pool, then choose the model to distribute credit inside it, and you document both so your reporting stays comparable over time.
It depends on how long your customers take to buy, so start from your sales cycle instead of a platform default. For low-consideration ecommerce where most purchases happen within a day or two, a 7-day click window captures nearly all the real conversions without crediting unrelated later activity. For considered B2B or high-ticket purchases that take weeks, a 14 or 30-day click window stops you from undercounting conversions that the ad genuinely influenced. In ChatGPT Ads specifically, the click window is configurable and pairs with a short fixed view-through window, so set the click length to your cycle and keep the view-through reporting separate. A practical approach is to pull the time-to-purchase distribution from your past orders, pick the window that covers the bulk of real conversions, and apply the same window across every paid channel so your comparisons hold. Revisit the choice only when your buying cycle or product mix actually changes.
Different platforms default to different window lengths, count clicks and views differently, and deduplicate conversions in their own ways, so the same sales rarely add up the same way twice. One platform might run a short click window with a brief view window while another runs a long click-only window, which alone produces two different totals from identical customer behavior. On top of that, each platform can only credit conversions it can match to its own click or impression, so cross-device and logged-out journeys fall into different buckets on each system. Reporting dates, modeled versus observed conversions, and the exact conversion event definition shift the numbers further. The fix is to compare like for like: read every platform on the same click window, use one shared conversion definition and value, and align the date ranges before you trust any gap between dashboards. Treat each platform total as the answer to its own narrow question; none of them is a universal truth.
Yes, most platforms let you change the attribution window on a live campaign, but doing it mid-flight quietly rewrites how your past and future conversions are counted. Because the window is a reporting rule applied on top of each sale instead of a fixed property of it, switching from a 7-day to a 30-day window can make the same campaign suddenly look stronger without anything changing in the market. That makes before-and-after comparisons unreliable, since you are no longer measuring the campaign on consistent terms. If you must change it, do so between clearly separated reporting periods, note the exact date you switched, and avoid drawing performance conclusions that straddle the change. The cleaner habit is to choose the window deliberately before launch, keep it stable through the campaign, and revisit it only when your buying cycle genuinely shifts. Stability in the window is what keeps your trend lines honest over the life of a campaign.
A good attribution window is the one that matches how your customers actually buy. Length alone, whether longest or shortest, is never the goal. If most of your conversions land within a few days of the ad, a short click window will capture nearly all the real credit while keeping noise out; if your buyers take weeks to decide, a longer window stops you from undercounting influence. The practical test is coverage: pick the window length that captures the bulk of your genuine time-to-purchase distribution, then confirm it credits a believable share of conversions without sweeping in activity your ads plausibly did not drive. A good setting is also consistent across every channel you compare and stable long enough to produce trustworthy trends. If different stakeholders argue over results, that usually signals mismatched windows instead of a bad channel, so agree on one standard and hold it.