Key takeaways
- Always rebuild the listing before driving paid traffic. Always.
- Run a manual-exact core, an auto/broad discovery layer, defensive brand campaigns, and conquesting where margin allows.
- Target ACoS is set per campaign by SKU margin, not by an industry average.
- Sponsored Brands and Sponsored Display are not optional for serious accounts.
Amazon PPC is the most measurable advertising channel in retail — and the easiest to lose money on. Most accounts we audit have the same three problems: a single auto campaign doing all the work, no defensive brand campaign, and a target ACoS set on instinct rather than margin. This pillar guide walks through how we actually structure and run profitable PPC for UK Amazon brands in 2026.
The four-layer campaign architecture
Healthy Amazon advertising is built in four layers. Each layer has a job; mixing the jobs is where waste comes from.
- 1Manual exact-match core — your researched, high-intent keywords with controlled bids and tight ACoS targets.
- 2Discovery layer — auto and broad-match campaigns whose only job is to surface new search terms for harvesting.
- 3Defensive brand — exact-match on your own brand terms; cheap, essential, blocks competitor conquesting on you.
- 4Competitor conquesting — Sponsored Display product targeting on competitor ASINs where your margin can absorb the higher ACoS.
Why the discovery layer matters
The exact-match core only works because the discovery layer keeps feeding it. Search-term reports from auto/broad campaigns are reviewed weekly; promoted converting terms graduate to manual exact, low performers get negative-keyword treatment.
Setting ACoS targets that actually protect margin
There is no industry-standard "good" ACoS. The right target is the one that leaves your unit economics intact after referral fees, FBA, returns, VAT and the actual product cost. We model break-even ACoS per SKU first, then set campaign targets at 60–80% of break-even for established ASINs and 100–130% of break-even during launch (where you are intentionally buying rank). Anything else is guessing.
Sponsored Brands and Sponsored Display — not optional
Sponsored Products gets the headlines, but a serious account uses all three formats. Sponsored Brands defends branded search, owns the top-of-results banner, and (with video) lifts CTR meaningfully. Sponsored Display drives both top-of-funnel awareness via audiences and bottom-of-funnel remarketing to shoppers who viewed but did not buy. Skipping these formats is leaving the most efficient sales on the table.
PPC during launch is its own discipline:Read the Amazon Product Launch pillar →
The weekly cadence that keeps PPC profitable
- 1Monday — review last week's search-term reports across all auto/broad campaigns. Harvest, negate, document.
- 2Wednesday — bid adjustments on manual exact campaigns based on rolling 14-day data. No reactive single-day moves.
- 3Friday — placement performance review (top-of-search, rest-of-search, product page). Adjust placement modifiers.
- 4Monthly — campaign restructure, dayparting review, share-of-voice check on category head terms.
Reporting that tells you what changed and why
Monthly PPC reports should answer three questions: did spend produce profitable sales, what changed about the structure this month, and what hypothesis are we testing next. We report spend, ACoS, TACoS, impressions, CTR, CVR, search-term harvest decisions and SKU-level ROAS — plus a written commentary that connects the numbers to the actions taken. Numbers alone are noise.
PPC sits inside the wider growth model:← Back to the Amazon Growth pillar
“When PPC is genuinely working, ad spend goes up and TACoS comes down. When it is not, both go up together. That is the whole reporting framework on one line.”
— Billy, Buy Box Savvy
