Use two windows, not one. Run and optimize Pinterest on 7-day click plus 1-day view, and hold it to a second, stricter floor: 7-day click-only in Ads Manager, or a third-party model that counts clicks plus verified views. Never judge Pinterest on 1-day click. That window cuts out the way Pinterest actually sells, and it is the fastest way to turn off a channel that is bringing you new customers.
Most arguments I see about whether “Pinterest works” are not about performance. They are about the window. Settle the window first and the performance conversation gets short.
What attribution windows does Pinterest offer?
Pinterest credits a conversion to an ad if it happens within a set number of days after someone saw, engaged with or clicked a Pin. Its Conversion insights help page lists three window types: view, engagement and click. If you never pick a window in Ads Manager, it falls back to 30-day click, 30-day engagement and 1-day view.
Two settings matter more than most teams realize:
- The window itself. How many days after a view, engagement or click a purchase still counts.
- The reporting date. Pinterest’s reporting API can report a conversion on the day of the ad interaction or on the day of the conversion. If your dashboard and Ads Manager use different settings, the same week will show different numbers.
When we connected a reporting tool to one fashion client’s Pinterest account this fall, the first pull didn’t match Ads Manager. It only matched once we set it to 7-day click, 1-day view, counted on conversion date. Check both settings before you compare any two reports.
Why does Pinterest look so different by window?
Pinterest is a planning platform. People save ideas days or weeks before they buy, and many never click the ad that put the product in their head. Pinterest has said that up to 95% of conversions its tag captures are view-based, and that 75% of conversions happen a week or more after ad exposure (Pinterest Business, 2022). Those are Pinterest’s own figures, so treat them as directional. But they match what we see in accounts.
In one premium personal care account, about 82% of Pinterest-reported checkouts over a 60-day stretch were view-through. Only about 1 in 5 came from a click. The brand’s third-party tool was set to 1-day click-only, so by design it counted none of the view-through sales. Pinterest showed roughly 3.5x ROAS. The third-party tool showed under 0.3x. Same ads, same customers, two very different stories.
The same account showed how much a short click window shortchanges Pinterest specifically. When we widened the view from 1-day to 3-day click, Pinterest’s ROAS rose about 80%. The brand’s other paid social channels rose about 20% to 35% on the same change. Pinterest buyers simply take longer to come back.
Which window should you use to judge Pinterest?
Here’s how the common options compare.
| Window | What it counts | Where it misleads | Best use |
|---|---|---|---|
| 1-day click | Purchases within 24 hours of a click | Misses most of Pinterest’s planning-cycle buyers and all view-through sales | Not recommended for Pinterest decisions |
| 7-day click only | Purchases within a week of a click | Still ignores people who saw the ad and bought later without clicking | Your conservative floor for scaling decisions |
| 7-day click + 1-day view | Clicks within a week plus views within a day | Can over-credit views on its own, especially for retargeting | Day-to-day optimization in Ads Manager |
| 30-day click + 1-day view (default) | Clicks within a month plus views within a day | Long click window inflates numbers versus other channels | Context only; rarely the headline number |
| Third-party clicks + verified views | Clicks, plus views matched to real orders | Needs setup and a supported integration | The cross-channel “real revenue” view |
My default is 7-day click plus 1-day view as the working number, with 7-day click-only as the floor. Neither number alone tells the truth. The pair does.
How do you set ROAS targets across two windows?
Set a target for each window, and only scale when both are met. Here is the process we use:
- Pick the two windows before launch. Write them down with the client or your finance lead. Agree that 1-day click is off the table for Pinterest.
- Match every report to those settings. Same window, same reporting date, in Ads Manager and in any dashboard or connector.
- Set a target for each. For example, a 5x on 7-day click plus 1-day view and at least 1x on 7-day click-only. Your margins set the real numbers.
- Read campaigns on both. A campaign that only looks good on the view-inclusive number is likely getting credit for people who would have bought anyway.
- Scale only when both hold. Then give it a few days after any big budget change before you judge it.
In a boutique fashion account we manage, that is exactly the rule we set going into Q4: keep scaling as long as the account holds 5x or better on 7-day click plus 1-day view and 1x or better on 7-day click-only. In September the account went from about 3.7x to 5.9x on the first and from about 0.95x to 1.4x on the second, so both conditions were met.
The two-window view also changed where the money went. One video campaign in that account showed the best view-inclusive ROAS, around 8x, but only about 0.6x on clicks alone. A catalog campaign showed lower view-inclusive numbers but was the strongest on click-only, at close to 1.9x. We kept the catalog campaign as the main driver of spend and held the video campaign to roughly 30%. If we had judged on one window, we would have funded the wrong campaign.
The same account also showed why patience matters. After a 2x budget increase, performance dipped for about three days, then recovered above goal, with click-only ROAS settling in a tighter 1.3x to 1.6x range. A one-window, three-day read would have pulled the budget back.
Should you trust Pinterest’s numbers or a third-party tool?
Use both, but set them up so they can actually agree. Platform reporting will always credit itself generously. A third-party tool set to clicks-only, or to a 1-day window, will always undercount Pinterest. The fix is a model that counts views you can verify.
Northbeam’s Clicks + Deterministic Views model is one example. It adds platform view data only when it can match a view to a real order through identifiers like order ID or hashed email. For Pinterest you have to turn on the Pinterest integration and select Pinterest in the model settings. That’s our recommended model for Pinterest when a brand uses Northbeam. Clicks-only is not ideal for Pinterest, and modeled views are less reliable than verified ones. Whatever tool you use, the principle is the same: credit views you can tie to an order, and compare channels on the same model.
The point is not to find the one “true” number. It’s to judge every channel by the same rules, so Pinterest isn’t held to a stricter standard than the channels it’s competing with for budget.
What else should you do to prove Pinterest is working?
- Track new-customer revenue. In the personal care account above, about 80% of the Pinterest revenue the third-party tool did count came from new customers. That’s the value a short window hides.
- Fix tracking before you argue about windows. If the Pinterest tag and Conversions API aren’t firing correctly, no window will be right. We cover what to check in How to Choose a Pinterest Ads Agency.
- Run a lift test when the stakes are high. Pinterest recommends conversion lift studies to measure incremental impact and suggests pairing test results with media mix modeling (Pinterest Business, 2024). A lift test answers the question every window argument is really about: would these sales have happened anyway?
- Compare like with like. If you are weighing Pinterest against other social channels, see Pinterest Ads vs Meta Ads for Ecommerce.
If you’re not sure your account is measured fairly, our Pinterest ads team includes former Pinterest employees and has run Pinterest ads since 2017. You can also start with the free Scorecard.
