The truth about CPCs — what the average hides.

The aCPC in your report is an average. Averages are useful for trend tracking and dangerous for tactical decisions. This episode unpacks what the average actually averages, and why two campaigns with the same aCPC can have wildly different economics.

The bimodal distribution

In almost every healthy Amazon PPC campaign, the actual CPC distribution has two humps:

• A cluster of low-CPC clicks from rest-of-search and product-page placements, where competition is light and Amazon's dynamic bidding lowers your bid.
• A cluster of high-CPC clicks from top-of-search, where competition is fierce and your placement modifier is in play.

The aCPC is the weighted average of the two. It almost never describes the experience of any single click.

What this means for bidding

If you tune the base bid to the aCPC, you systematically misprice both clusters. The cheap cluster gets paid more than it would on its own; the expensive cluster falls below its competitive threshold and stops winning. The result is a campaign that maintains aCPC while losing impression share at the slot it wanted most.

Where CPC drift comes from

• Auction entry by new bidders. A new competitor entering the auction raises the next-highest-bid that you have to clear.
• Seasonal demand spikes. Q4 raises CPCs across the board for 6–10 weeks.
• Your own placement modifiers. If you push Top-of-Search to +200 %, you have voluntarily raised your top-of-search CPC ceiling.
• Conversion-rate decay on your ASIN. If CVR falls, Amazon's expected-revenue calculation drops, so to keep impression share you end up paying more.

The practical rule

Never bid against the aCPC. Bid against the placement-level CPC distribution for the slot you actually want to win, priced from your goal-derived target bid. The aCPC is for monthly health tracking, not for tactical bid changes.