Amazon PPC Strategy: Structure, Bidding, and the Weekly Loop That Makes It Work

What an Amazon PPC strategy is actually built from

Strategy isn't a bid number. It's a structure. Each campaign should have one job, a bidding rule that fits that job, and a review cadence that acts on real search-term data rather than a hunch.

The structure usually has four layers. An auto campaign finds the search terms Amazon thinks are relevant — its job is discovery, not profit. A manual exact campaign holds the terms you've already proven convert, with a bid you control. A defensive branded campaign protects your own listing from competitors bidding on your product name. And a category or complementary campaign puts your product in front of shoppers who haven't searched for it yet — awareness spend, not conversion spend.

If you sell through FBA, add one more input: inventory. An amazon fba ppc strategy that ignores days-of-cover will happily push bids up on a listing that's a week from a stockout. The bid math doesn't know that. You have to check it manually, or build a rule that does.

The bidding strategy: pick the rule before you pick the number

Amazon gives you three ways to let its bid algorithm move your bid at auction time, and choosing between them matters more than the bid figure you type in. See the table below for the tradeoffs of each.

Whichever mode you choose, check every bid against one number: your breakeven ACoS. Say a listing sells at $28.99 with a 40% margin — $11.60 per unit. Breakeven ACoS is margin divided by price, so 40%. Anything you spend below that on a converting term is profit; anything above it is a loss you're choosing to accept, usually for rank or launch reasons. That number — yours, calculated from your own margin — is what every bid decision should be checked against, not a round figure like '20% ACoS' pulled from someone else's blog post.

The weekly loop: harvest, promote, negate — with a rollback trigger

The loop is harvest, promote, negate, and it runs weekly once a campaign has enough data — usually 14 days or 15 to 20 clicks per search term, whichever comes later. Deciding on less than that is deciding on noise.

Worked example: your auto campaign's search term report shows a term with 22 clicks, $19.80 spent, 3 orders, and $86.97 in revenue. ACoS is 22.8% — below the 40% breakeven above, so it's a keeper. Move it into a manual exact campaign with a bid you control, and add it as a negative exact in the auto campaign so the two stop competing for the same click. The same report shows a second term with 31 clicks, $27.90 spent, and zero orders. Past the 20-click threshold with nothing to show for it, that one becomes a negative keyword everywhere.

Whatever you change, write down three things before you touch the bid: the evidence that justified it, how you'll measure whether it worked, and the number that tells you to reverse it. That third piece — the rollback trigger — is the step most sellers skip, and it's the reason so many 'optimizations' quietly make things worse for a month before anyone notices.

Budget by funnel stage, and what to do when the ACoS is wrong

One common starting split, worth treating as a first guess rather than a target: roughly 40% to discovery, 35% to proven exact-match converters, 15% to defensive branded, 10% to category and awareness. Adjust it weekly against what the search-term report actually shows, not against the split itself.

When account-wide ACoS is running well above your breakeven — not one campaign, the whole account — check three things before you touch a single bid. Did a competitor enter the auction and raise the going rate? Did price change without the ACoS math following it? Did a listing go out of stock and buyers scatter to whoever was left? Bid changes fix targeting problems. They don't fix a stockout, a price change, or a new competitor, and moving bids to solve those just makes the real problem harder to see later.

The mistakes worth naming, including ones we've made

• Harvesting too early. Promoting a term to exact match after 3 clicks and one order builds a strategy on a coin flip.
• Negating during launch. A brand-new listing needs volume for Amazon's algorithm to learn who converts. Cutting terms in week one to protect ACoS can starve that learning before it starts.
• Chasing ACoS to zero. The lowest-ACoS version of a campaign is usually also the lowest-volume version. Below breakeven, cheaper isn't always better — it can mean invisible.
• Treating a stockout like a targeting problem. We've done this ourselves: pulled bids back after one bad week that turned out to be a supply issue, not a keyword issue, and spent the next week recovering rank from a bid cut that never needed to happen.

Doing this by hand, or automating the same loop

Everything above is doable manually. It's a spreadsheet, a search-term report pull, and a recurring hour on the calendar every week — the ceiling is usually time, not knowledge.

Where it stops scaling is account level: running this loop correctly across dozens of SKUs, multiple marketplaces, and daily inventory shifts is running dozens of spreadsheets at once. That's the gap Dr. PPC (drppc.ai) is built for — it reads the account, writes a strategy per product against that brand's real economics, and proposes each change with the evidence behind it, a measurement plan, and a rollback trigger, the same three things described above. Full Circle, which operates it, has managed more than $500M in revenue across 100+ brands, and the software behind it, Orbit, is included rather than billed separately. Whether or not you ever use it, the loop underneath is the one you'd run by hand.