The claim
“Impressions are down. Just bump the bid 20% and they'll come back.”
It is the most common fix I see, and the most common one I used to reach for. It feels like turning up a dial.
The usual reading
The reading treats the auction as one step: everyone who wants the search shows up, the highest effective bid wins, so more bid means more wins means more impressions. If impressions are low, bid is the lever.
Premises, laid out
Every conclusion below rests on these. Reject one and you can see which conclusions fall with it.
- ❌Every search you care about is an auction you are already in. Before the auction there is a gate: whether your product is eligible to be matched to that shopper's intent at all. I think of the keyword as an entry ticket, not the match itself. Bid does not issue tickets.
- ✅Bid affects whether you win among eligible competitors. It does. The question is how much, and what it costs.
- ❌Your campaigns compete independently. When several of your own campaigns target the same query, Amazon lets only one of them into that auction (this is in Amazon's own targeting guidance). Raising one campaign's bid can simply move volume between your own campaigns.
- ⚠️Budget isn't the binding constraint. For many campaigns it is.
How we take it apart
Free: conclusion + one screenConclusion: impressions are produced by three stages — eligibility, ranking among eligible ads, and then shopper response — and the bid only acts on the middle one. Before touching the bid, check whether the campaign is budget-capped and whether a sibling campaign is taking the auction.
The full step-by-step decomposition, including how AI joins the reports, is in the course.
Evidence
Free: one exampleExperience
A bid step on one product. I cut the bid on one product's two main campaigns by 14.3%. Read at the campaign level, search impressions fell 25.4% — it looked like a failure. Read at the product level, they fell 7.9%, because about 65% of the lost volume was picked up by my own other campaigns on the same product. Bid changes have to be read at the product level, not the campaign level.
What would prove this wrong
Free: one sentenceDisproof condition. If, in campaigns that are not budget-capped and have no sibling campaign on the same query, a controlled bid increase reliably produced proportional impression gains at the product level, I would move this to ⚠️.
How you run it
Course onlyReproduce this on your own account
- Step-by-step operations, including the AI prompts used to join reports
- Query templates and the exact fields to pull
- A checklist to rule out the common reading traps
- The full disproof conditions and the failed cases
Ledger status
Verdict: ❌ doesn't hold — as a general rule. Bid buys a higher win rate among eligible ads; it does not buy eligibility, it can be cancelled out by your own campaigns, and in capped campaigns it trades win rate for fewer chances.
Type, in detail: Experience (own accounts).
Have data that says otherwise? Share a case. Reversals are published.