Retargeting ROAS collapses after a specific pool-age threshold because the ratio of fresh visitors entering the pool to stale visitors leaving it drops below roughly one-to-three. When that happens, the retargeting ad set is showing the same ads to the same people who already decided not to buy, and ROAS falls off a cliff within two weeks regardless of creative quality. The fix is almost never creative. The fix is pool composition — specifically, the inflow rate of new visitors into the retargeting window. Most retargeting ad sets that stop working stop working because the top-of-funnel feeding them slowed down, not because the ads got worse.
Retargeting pools decay when the fresh-visitor inflow drops below 1:3 of the stale-outflow rate. The failure is composition, not creative. Creative refreshes buy weeks. Pool composition fixes — scaling prospecting or extending the window — solve the structural cause. Audit the ratio monthly, act when it crosses the threshold.
What is retargeting pool decay in plain terms?
Retargeting pool decay is the state where a retargeting ad set's audience is composed mostly of people who already saw the ads, considered the product, and chose not to buy. The pool is still technically populated — the thirty-day website-visitor audience still contains the right number of people. But the composition has shifted. The fresh visitors who entered yesterday are a small fraction. The majority were there last week and the week before, and they have seen the same ads multiple times already. The retargeting system is now paying to show ads to a declining-value audience that has already made its decision. ROAS lags the composition shift by one to two weeks because the attribution window is still catching conversions from fresh visitors who entered earlier.
What is the specific threshold that triggers pool collapse?
The threshold is roughly one-to-three: for every one fresh visitor entering the retargeting window per day, no more than three stale visitors should remain in the active pool. Expressed differently: at least twenty-five percent of the retargeting pool should be visitors who entered in the last seven days. When fresh entries fall below this ratio, the retargeting ad set is predominantly serving ads to a decided-no audience, and CPA begins climbing within one week. Within two to three weeks, ROAS reflects the shift. The threshold is derived from the observation that visitors who did not convert within seven days of their initial site visit have a decision-reversal rate below five percent in most DTC categories, so once they dominate the pool composition, the retargeting ad set is spending against the wrong audience.
Why do creative refreshes not fix the decay?
Creative refreshes do not fix pool decay because the failure is not that the audience is bored with the ads. The failure is that the audience already decided. A new creative angle on the same decided-no audience produces maybe a week of lift as the novelty refreshes attention, then the same underlying rejection reasserts itself. The audience saw the product, considered the price, evaluated the fit, and chose not to buy. A new photo or hook does not re-open the decision. Pool decay follows the same decay pattern that audience saturation shows, and the practical implication is that retargeting creative refreshes should be measured against pool composition, not against ad fatigue. If the pool is still fresh and ads are tired, refresh works. If the pool itself has aged, no creative recovers it.
What are the signs your retargeting pool is decaying before ROAS drops?
1) Frequency on the retargeting ad set rising above four within fourteen days of launch. 2) Fresh website visitors (last seven days) representing below twenty-five percent of the thirty-day retargeting audience. 3) Prospecting campaign spend dropping or CPM rising while retargeting spend holds steady. 4) Click-through rate on retargeting creative holding stable but conversion rate declining — the audience is still clicking out of habit but no longer converting. 5) Return visitor rate from the retargeting ad set above sixty percent (same people, same clicks, no new sales). Any three of these in combination is a pool composition problem, not a creative problem.
How do you actually fix pool decay structurally?
The structural fix is to restore inflow. Three levers: scale prospecting spend by fifteen to twenty-five percent to feed more fresh visitors into the retargeting window; extend the retargeting window from seven or fourteen days to thirty or sixty days to pull in a longer tail of visitors (this works best for considered purchases where the research window is naturally longer); or add new fresh-traffic sources (organic, email, Google) that increase pool inflow without increasing Meta prospecting cost. Prospecting efficiency determines retargeting health in a way most account structures do not budget for, which is why the retargeting conversation almost always points back to how much fresh traffic the top of funnel is producing.
When should you retire a retargeting ad set entirely?
Retire the retargeting ad set when the inflow-to-outflow ratio falls below one-to-five and cannot be restored because prospecting cannot be scaled further and the window is already at sixty days. At that point the ad set is spending efficiently against an audience that has structurally decided. The practical move is to pause the ad set for thirty days to let the pool refresh naturally through continued site traffic, then relaunch. In Advize's experience, retargeting ad sets paused and relaunched after a thirty-day cooling period often outperform their original run because the pool composition has been restored through organic and direct traffic in the interim.
Conclusion
Retargeting pool decay is a composition problem disguised as a creative problem, which is why most teams try to fix it with new ads and watch the fix fail. The useful intervention is restoring the inflow-to-outflow ratio: either by increasing the fresh-visitor rate from prospecting and organic, or by extending the retargeting window to pull in a longer tail of visitors. Advize's practical rule is to audit the ratio monthly and treat anything below 1:3 as a strong signal that the retargeting ad set has weeks, not months, before ROAS reflects the decay.