How attribution windows credit the click

Click versus view#

Click attribution credits conversions to the ad someone tapped; view attribution credits conversions to an ad someone merely saw. Clicks signal intent, views signal exposure, and the two behave nothing alike: view-credited conversions flatter prospecting while click-credited ones discipline it. Know which your reports blend before judging any campaign. Start at the Meta Ads hub for the family view.

Default reporting mixes both, which is why two advertisers can quote different ROAS from identical delivery. Split click and view columns in every report that informs budget, and judge prospecting on clicks first. Meta own attribution windows guide lists the current options.

The 7-day click standard#

Seven-day click plus one-day view is the default lens for most accounts: a week for considered clicks to convert, a day for views to claim stragglers. Longer click windows suit considered purchases with research phases; shorter windows suit impulse offers where delayed credit mostly adds noise. Match the window to the buying cycle, not to vanity.

Changing windows rewrites history cosmetically: past delivery is untouched, but reported totals shift, which confuses stakeholders mid-quarter. Announce window changes like methodology changes, with before-and-after notes. Learning phase maths assumes the window you actually run.

How windows move reported ROAS#

Wider windows report more conversions against the same spend, so ROAS rises without a single extra sale. Narrower windows do the reverse, and neither movement means performance changed. Judge creative and audience tests inside one fixed window; cross-window comparisons measure the ruler, not the work.

View windows move numbers most on video-heavy prospecting, where impressions are cheap and audiences are vast. If view-through share dominates reported revenue, the window is doing the selling, not the creative. Segment view credit out and re-read before scaling. Video reporting needs the split most.

Windows and learning volume#

Optimisation feeds on counted events, so tight windows starve learning accounts: a 1-day click window on a considered purchase may never clear roughly 50 events a week, while 7-day click gets there comfortably. Volume-hungry accounts should widen the click window before widening the audience. Consolidation compounds the effect.

Balance honesty against learnability: the window must be wide enough to train delivery yet tight enough to credit plausibly. Document the chosen window beside the budget logic, so future audits know why that ruler was picked. Developer event timing detail is at event parameters.

When to change windows#

Change windows when the buying cycle says so: longer consideration earns longer click credit, subscription renewals earn their own logic, and flash sales earn tighter reads. Never change windows to rescue a failing narrative mid-month; that is methodology shopping, and finance teams spot it instantly.

Re-baseline targets after any change, because CPA and ROAS targets calibrated to one window misfire under another. Run old and new columns side by side for a fortnight so stakeholders see the translation. Our mismatch guide reconciles platform and analytics counts.

Comparing across platforms#

Every platform credits itself generously under its own windows, so summed platform totals always exceed order-book truth. Compare platforms on order-book reconciliation and incrementality holds, never on native ROAS side by side. Windows explain the gap; they do not excuse double counting the same sale twice.

Hold a monthly measurement review: window settings, overlap estimates, modelled shares and finance reconciliation in one sitting. Modelled conversions add a further estimated layer, and our team runs the review when internal debate stalls.

Frequently asked questions#