Most Amazon sellers set bids by feel: raise them when sales are slow, cut them when ACoS looks high. That works until it doesn’t. A bid is a price you’re agreeing to pay for a click, and the right price comes from your margin, not from a hunch.
This guide is for brands running Amazon ads that want a repeatable way to set bids and budgets. It covers the formula we start from, how placements and budgets change the picture, and the factors that should move a bid up or down each week.
The short version
- Start every bid from your margin: max cost per click = price × target ACoS × conversion rate.
- New keywords need data before they get a final bid. Start lower and let the clicks come in.
- A campaign that runs out of budget at noon is invisible for the rest of the day. Fix budgets before bids.
- Bid less where you already rank organically, and less when stock runs low. Bid more where clicks turn into revenue.
In this guide
- Start From Your Margin
- Bidding on New Keywords
- Placements: Where Your Ad Shows Matters
- Budgets That Last All Day
- Bid Around Inventory and Organic Rank
- Bid by Campaign Purpose
- Amazon’s Bidding Options
- A Weekly 30-Minute Bid Review
- FAQ
1. Start From Your Margin
Every bid decision starts with one number: how much you can afford to pay for an ad-driven sale. That’s your target ACoS, which comes from your break-even ACoS. We explain both in What Is a Good ACoS on Amazon?, but in short:
Break-even ACoS = (price – product cost – referral fee – FBA fee) ÷ price
Target ACoS = break-even ACoS × 50% (profit first), 75% (balanced) or 90% (growth)
From there, the maximum you can pay for a click follows:
Max cost per click = price × target ACoS × conversion rate
| Price | Target ACoS | Conversion rate | Max cost per click |
|---|---|---|---|
| $30 | 30% | 12% | $1.08 |
| $30 | 30% | 6% | $0.54 |
| $50 | 25% | 10% | $1.25 |
| $20 | 35% | 15% | $1.05 |
The second row is the one to stare at. Same product, same target, but half the conversion rate means half the affordable bid. That’s why a weak listing makes ads expensive: you can’t outbid a conversion problem.
The same formula works per keyword. Use each keyword’s own conversion rate once it has enough clicks, and your product’s overall rate until then.
2. Bidding on New Keywords
A new keyword has no conversion data, so its true value is unknown. Two mistakes are common: bidding high to “get data fast”, and cutting the keyword before it’s had a fair chance.
What we do instead:
- Start lower than the formula suggests. Bid towards roughly half of your break-even ACoS until the keyword has conversions to judge.
- Give match types different patience. Exact keywords give a clear signal quickly. Phrase and broad keywords match many searches, so they need more clicks before you judge them: in our own bid rules, broad gets about twice the clicks of exact before its bid starts coming down.
- Lower the bid gradually as clicks pile up without a sale, rather than cutting it at a fixed number. A keyword that hasn’t converted after 5 clicks deserves a smaller bid, not a pause.
- Look at the lookback window. Judge exact keywords on about 14 days of data and broad on about 21, so slow converters aren’t cut too early.
3. Placements: Where Your Ad Shows Matters
Sponsored Products ads show in three places, and each converts differently:
| Placement | Where | What usually happens |
|---|---|---|
| Top of search | The first row of results | Most expensive, usually the best conversion |
| Rest of search | Further down the results | Cheaper, middling conversion |
| Product pages | On other products’ detail pages | Cheapest clicks, often the weakest conversion |
Amazon lets you raise bids by placement. We set those adjustments from revenue per click (ad sales ÷ clicks). The placement with the lowest revenue per click is the baseline. Every other placement gets a boost in proportion to how much more revenue each click brings in.
For example, if top of search brings $5 per click and product pages bring $1, top of search can justify a much higher bid, because each click is worth five times as much.
An apparel brand we worked with had more than 200 products lumped together in the same campaigns. Splitting out the top products and putting spend where it converted lifted the brand to its best month ever, $578K in revenue, at an ACoS just under 11%, while holding its 10% TACoS goal. Read the case study
4. Budgets That Last All Day
A campaign that runs out of budget stops showing until midnight. If that happens every day at noon, you’re invisible for half of every day, usually including the evening, when many shoppers buy.
A supplements brand we took over had exactly this problem: budget limits caused campaigns to pause for a few hours every day, which killed momentum in the rankings. Fixing that was one of the first changes. The brand went on to grow 15x from $42K/Month in monthly sales. Read the case study
How to set budgets:
- Check budget usage first. If a profitable campaign runs out of budget most days, raise the budget before touching bids.
- Separate proven from experimental. Give exact-match campaigns on proven keywords their own budget, so discovery campaigns can’t drain it.
- Lower bids rather than capping budgets on campaigns that spend too much. A capped campaign spends its money on the morning’s shoppers, not the best ones.
- Know your testing cost. Ten new keywords at about $2 a click, each needing at least 10 clicks to test, is $200 before you learn anything. Budget for it.
Not sure your bids match your margin? We’ll look at your account and show you where the margin is going. You decide what to do with it. Get my free audit
5. Bid Around Inventory and Organic Rank
Two things outside the ad account should change your bids: how much stock you have, and where you already rank.
Inventory. Running out of stock hurts organic rank, and spending ad money to sell your last units faster makes it worse. As days of stock fall, we pull bids back in steps:
| Days of stock left | Bid change |
|---|---|
| Under 7 | Cut sharply, about 75% |
| 7–14 | Cut about 50% |
| 14–21 | Cut about 25% |
| 21–30 | Cut about 10% |
| 30–60 | No change |
| 60–90 | Raise about 10% |
| Over 90 | Raise about 20% |
Calculate days of stock from total sales, organic and ad, not ad orders alone, or you’ll think you have far more time than you do.
A pet brand we worked with ran exactly this kind of inventory-aware bidding, with strict profit guardrails so bids and budgets never ran at a loss. It grew from £124K/Month to £434K/Month in 10 months. Read the case study A mattress brand ran ad spend against its stock the same way, reaching $224K/Month with TACoS under 14%. Read the case study
Organic rank. Don’t pay full price for visibility you already have. If you rank in the top three organically for a search, you’re already at the top of the page, so we cut the top-of-search bid for that keyword sharply and keep product-page spend, where you’re still competing against other products. If you’ve slipped to page two, bids go up to win the rank back.
6. Bid by Campaign Purpose
Not every campaign has the same job, so not every campaign should be judged on the same ACoS.
| Campaign purpose | What it’s for | How to bid |
|---|---|---|
| Profit | Proven searches that sell | Profit-first target, tight control |
| Discovery | Finding new searches | Low bids, broad and automatic, judged on what it finds |
| Ranking | Pushing a product up for key searches | Higher bids for a set period, judged on rank and total sales |
| Defense | Your brand name and your own product pages | Low bids, high conversion, keeps competitors out |
For launches, we don’t build everything around expensive exact-match bids at the top of search. We put more budget into broader targeting and let the data show which searches deserve their own campaigns. How we push a product up for one competitive search is in Amazon Ranking Strategy 2026, and the campaign types we use most are in Top 5 Amazon PPC Campaigns That Are Actually Working in 2026.
As accounts grow, campaign counts grow with them. A popcorn brand we worked with went from 15 to 144 active campaigns as it scaled, peaking at $112K in one month. Read the case study
7. Amazon’s Bidding Options
Amazon offers three bidding strategies for Sponsored Products campaigns:
- Dynamic bids – down only. Amazon lowers your bid when a click looks less likely to convert. The safest default.
- Dynamic bids – up and down. Amazon can raise your bid (by up to 100% at top of search) when a click looks more likely to convert. Use it on proven campaigns, not new ones.
- Fixed bids. Amazon uses your exact bid. Useful for tests where you need clean data.
Amazon also offers rule-based bidding, which adjusts bids automatically toward a ROAS target. We tested it across about 30 brands and most campaigns improved, but it needs data to work with: at least 10 conversions before the rules have anything to go on. Don’t use it for launches or new campaigns.
Whatever tool changes your bids, cap how much any single change can move them. We limit each adjustment to about 25% up or down, so one bad day of data can’t swing a bid too far.
8. A Weekly 30-Minute Bid Review
Once a week, with at least 14 days of data:
- Budgets (5 minutes). Which profitable campaigns ran out of budget? Raise them.
- Overspenders (10 minutes). Which keywords are well above target ACoS? Lower their bids, by no more than 25%.
- Underspenders (5 minutes). Which keywords are well below target but getting few impressions? Raise their bids.
- Stock (5 minutes). Which products are under 30 days of supply? Pull their bids back.
- Rank (5 minutes). Which main keywords moved in organic rank? Adjust top-of-search bids to match.
Search terms and negatives are a separate, every-other-week job, covered in our Amazon PPC targeting guide.
Not sure your bids match your margin? We’ll look at your account and show you where the margin is going. You decide what to do with it. Get my free audit
9. FAQ
Use max cost per click = price × target ACoS × conversion rate. A $30 product with a 30% target ACoS and a 12% conversion rate can afford about $1.08 a click.
Enough that your profitable campaigns don’t run out of budget, plus a set testing budget for new keywords. Judge the total on TACoS: ad spend as a share of all your sales should stay below your margin.
On proven campaigns, yes. On new campaigns or launches, start with down only, so Amazon can’t double your bids before you know what a click is worth.
Weekly for most accounts, using at least 14 days of data, and by no more than about 25% at a time.
Extra bid percentages for top of search, rest of search and product pages. Set them from revenue per click: the placement that earns the most per click gets the biggest increase.
Usually because bids are too low to win placements, the budget is too small, or the targeting is too narrow. Check impressions first: few impressions means bids or targeting, not budget.
Related Guides
- Amazon Seller Fees Explained: Every Fee and Your Real Margin
- 12 Amazon PPC Mistakes We Find in Almost Every Account Audit
- Amazon PPC Match Types: Broad, Phrase, Exact and the 4th Type
- The Amazon PPC Tools We Actually Use
Chris Rawlings is the founder of Sophie Society and an Amazon Ads Verified Partner. Sophie Society manages $5M/Month in Amazon ad spend and has worked with 800+ brands.
