What is conversion lag (ibnr)?
Conversion lag is the delay between an ad click and the conversion event Google eventually reports for it, which means any recent date window is undercounted at the moment you look at it and gets backfilled for weeks afterward. IBNR (incurred but not reported, a term borrowed from insurance) describes exactly this: real conversions that already happened but haven't been reported yet, so a raw recent-window count always understates the true number.
Google reports a conversion against the date of the original click or interaction, not the date the conversion actually happened, and it keeps backfilling that count for weeks as delayed conversions settle in. That means a campaign's last 7 days of reported conversions, viewed today, is genuinely incomplete, not wrong exactly, but a partial count that will keep rising.
The practical rule this demands: never treat a recent window's raw conversion count as final without a maturity or completeness estimate alongside it, projecting forward from the account or campaign's own historical settling curve. A safe threshold worth adopting: only call a recent window's numbers 'hard data' once its estimated completeness crosses roughly 90%, below that, report both the raw count and the projected final count together, and label the raw number as directional, not a verdict.
This especially matters for pause/kill decisions. 'Zero conversions this week, pause it' is a dangerous claim on an immature window with low traffic, since a genuinely working entity can show zero simply because its conversions haven't settled yet. The safer bar combines both a maturity threshold and a minimum traffic floor (enough clicks that a true zero would be a meaningful signal) before treating a dark window as a real verdict rather than noise. Note this applies specifically to Google's click-attributed conversions; Meta uses a different attribution basis without the same lag curve.