Static click thresholds cost real money
Almost every Amazon agency I’ve seen uses the same heuristic for cutting non-converting targets. If a keyword or target accumulates 20 clicks without an attributed order, pause it.
The rule sounds reasonable. It scales easily across clients. It can be operationalized by a junior account manager. It produces a defensible answer.
It’s also almost always wrong.
Here’s why.
The right click threshold for “this target is not going to convert” depends on the account’s actual conversion behavior. Specifically: the account’s blended conversion rate, the variability around that rate, and the AOV of the products being advertised.
A pet supplement account I audited had a blended conversion rate around 4%. The math of statistical significance on a 4% CVR says you can be reasonably confident a target won’t convert after about 7 clicks with zero orders. The agency was using 20. Every non-converting target was wasting roughly three times the necessary spend before getting cut.
A baby monitor brand had a much lower conversion rate and a higher AOV. The right dynamic threshold there was 25 clicks. Cutting too early would have killed targets that needed more time to validate.
A premium body pillow brand had a $189 AOV and a meaningfully lower CVR than typical. The right threshold for them was 45 clicks. The generic 20-click rule would have killed half the campaigns before they had time to learn.
Three accounts. Three different thresholds. One agency rule applied to all three would have wasted money in three different ways.
What the dynamic threshold actually does
The math is straightforward.
Take the account’s blended conversion rate. Compute the click volume at which the probability of zero orders is statistically low enough to call the target dead. Adjust for AOV (higher AOV products need longer to validate because the trigger to buy is harder to pull).
For an account with a 2% CVR, the dynamic threshold is typically 12 to 15 clicks. For a 5% CVR account, it’s 7 to 10. For a 1% account with high consideration products, it can be 25 to 40.
Most agencies don’t compute this. They use 20 because 20 is what was in the standard SOP they inherited from the last agency the account manager worked at. Twenty has nothing to do with the math of the account they’re managing today.
What it costs
The cost scales with account size.
In one premium body pillow account, nearly 50% of total ad spend (about $14,700 in a 60-day window) was sitting in non-converting targets that had crossed the dynamic threshold but not the static one.
In a baby monitor brand at scale, the equivalent pool was over $81,000 in 64 days, 38% of all spend.
In a home textiles brand, the no-order pool was projected at over $300,000 over 90 days at current run-rate.
Three accounts. Combined exposure: nearly $400K in 90-day waste sitting inside non-converting traffic that the existing agency rules wouldn’t catch.
How to find it in your own account
Two steps.
First, pull your search-term report for the last 60 days. Filter to targets with more than 10 clicks and zero attributed orders. This is the candidate pool.
Second, ask your agency to compute the dynamic threshold for your specific account. If they can’t, or if the answer is “we use 20,” you know the diagnostic isn’t happening.
The fix isn’t fancy. It’s a weekly review of high-click no-order targets at the correct threshold for your account, with negative keywords or bid reductions applied to the validated waste.
Static rules scale. Static rules also cost money. Pick one.