KDP Ads Not Delivering? Why Your Campaign Gets No Impressions, and the Fix Ladder

You built the campaign, picked the keywords, set the budget — and three days later the console shows a row of zeros. No impressions, no clicks, no spend. Nothing is broken: an ad with no impressions is an ad that is losing every auction it enters, or entering none. Both have causes you can check in order, from most likely to least. Work the ladder top-down and stop at the first rung that changes the numbers.

First: is it actually a problem yet?

New campaigns and freshly raised bids do not deliver instantly. Amazon takes time to evaluate a new campaign before serving it at volume — 48–72 hours of thin delivery is normal ramp-up, not failure. Editing the campaign daily during this window restarts the evaluation and makes it worse. If it has been less than three days: wait. Genuinely.

Rung 1: the bid (the cause ~70% of the time)

Ad slots are auctions, and a bid below the going rate simply never wins. Beginners routinely bid $0.20 in niches where clicks clear at $0.80.

The test is mechanical: double the bid on a handful of keywords and watch 48 hours. Impressions appear — it was the bid; now find the lowest bid that still delivers. Still zero — move down the ladder. (Bidding higher does not commit you to losing money: your ceiling is your break-even ACOS, and if profitable delivery is impossible at any winnable bid, that keyword was never yours.)

Rung 2: relevance — will Amazon show your book for this at all?

Amazon only serves ads it expects to convert; a keyword unrelated to what your book's page says it is gets throttled regardless of bid. Classic version: bidding on a bestseller's name in a different genre. If your listing's title, categories and keywords say "cozy mystery" and you target "space opera," no bid fixes that. The fix lives in the listing, not the campaign — the same visibility work that fixes organic no-sales.

Rung 3: the "low search volume" trap

A keyword flagged low-volume is throttled because almost nobody types it. Hyper-specific long-tails ("clean small town enemies to lovers novella under 200 pages") feel smart and serve nothing. Replace them with the broader head terms readers actually use; let the search-term report tell you the long tail later. This is also the argument for keeping an automatic campaign alive: Amazon's own matching finds serving inventory your keyword list missed — the two-campaign structure exists precisely for this.

Rung 4: structural checks (fast, occasionally the answer)

• Campaign paused / out of budget? A $2 daily budget can exhaust on two clicks and read like "no delivery" for the rest of the day.
• Book eligibility. Ineligible or recently-changed listings (blocked category, in review) do not serve.
• Negative keywords. A broad negative added carelessly can block your own targets — the classic silent kill.
• All eggs in exact match. Exact-only campaigns with tiny keyword lists have tiny audiences by construction; widen with phrase match or more terms.

The opposite failure

Know what you are steering toward: a campaign that delivers too easily at a high bid can spend fast on clicks that never convert. Delivery is not the goal; profitable delivery is. Once impressions flow, judgment switches instruments — from the impressions column to ACOS versus your break-even, and TACOS on real royalties. Getting the ad served is plumbing. What you serve it for is the business.

Where TrueRoyalties fits

Fixing delivery means raising bids, and raising bids means your profit margin is now doing the tightrope walk. TrueRoyalties shows what each book actually earns per sale — royalties net of printing, tax and currency, KENP included — joined to what its ads spend, so when the impressions finally arrive you know immediately whether they are buying profit or just motion: Net Profit = Royalties − Ad Spend, per book, continuously.