Amazon bid types and strategies come down to three campaign-level settings and one set of
adjustments. Dynamic bids down only, dynamic bids up and down, and fixed bids, plus placement
adjustments that multiply your bid depending on where the ad appears.
The bid you enter is a maximum, not a price. Amazon runs a second-price style auction, so you
generally pay slightly more than the next competing bid rather than your full bid. That single fact
should change how you think about the number: raising a bid does not automatically raise your cost per
click, it raises your eligibility to win placements.
Understanding that is worth more than any recommended setting, because it explains why cautious bidding
frequently produces both fewer sales and no saving.
The three bidding strategies
Dynamic bids, down only
Amazon lowers your bid in real time when a click looks less likely to convert, and never raises it. In my experience that asymmetry is why it suits beginners.
The safest setting and the right default for most new campaigns. You cannot overspend relative to
your stated bid, and the worst case is missing impressions you might have won.
Use it for: new campaigns with no data, discovery campaigns, tight budgets, and any situation where you
do not yet know what converts.
Dynamic bids, up and down
Amazon lowers the bid when conversion looks unlikely and raises it, potentially substantially, when
conversion looks likely.
Powerful, and in our experience genuinely capable of surprising you. The raise applies to placements where Amazon
predicts a sale, which is usually top of search, and that is also the most expensive real estate.
Use it for: proven exact-match keywords where you know the conversion rate and the margin supports a
higher cost per click. I’d not put this on a campaign you have not watched for a few weeks, and I’d
set the base bid lower than you would under down-only, because the ceiling is higher than the number you
typed.
Fixed bids
Amazon uses your exact bid every time, no adjustment either way. Predictable, and I’d value that more than people do.
Useful for a narrow set of jobs, and I’d say over-used generally. It is the right choice when you are
deliberately gathering clean data, because nothing is being modified underneath you, and when defending
a brand term where you want predictable presence.
Use it for: brand defense, controlled tests, and situations where consistency matters more than
efficiency.
Placement adjustments
Separate from strategy and frequently ignored, which in our experience is a mistake because they multiply rather than add.
Amazon lets you increase your bid by a percentage for top of search (first page) and product
pages, with rest of search taking the base bid.
Top of search converts better and costs more. Not a contradiction. The whole trade. When I have
tested adjustments on keywords already proven to convert, a modest one has been the highest-return
change available in a mature campaign, and it is routinely left at zero.
Two cautions worth stating. Adjustments compound with dynamic up-and-down bidding, so combining an
aggressive placement adjustment with up-and-down on a high base bid can produce a cost per click well
above what you intended. And placement adjustments sit at campaign level, which is another reason
campaigns should be coherent units rather than mixed bags. Our guide to
Amazon PPC campaign structure covers why that
matters.
How to actually set the number
The part guides usually skip in favor of listing the options.
Start from your break-even, not from Amazon’s suggestion. Work out the maximum cost per click your
margin and conversion rate support: if your contribution margin per unit is $8 and the keyword converts
at 10%, then 10 clicks produce one sale worth $8, so anything above roughly $0.80 per click loses money
before you account for returns.
Then decide whether you are buying profit or buying data. Those are different jobs and they justify
different bids. A launch campaign deliberately buying visibility can run above break-even for a defined
period. A mature campaign should not.
Use Amazon’s suggested bid as a range indicator, not a target. It tells you roughly what the auction
looks like. It knows nothing about your margin.
I’d write the break-even figure down per product before touching any campaign. In our experience most
bidding problems are actually margin problems that nobody calculated.
What to change, and in what order
Bids are the most adjustable thing in an account, which makes them the most fiddled with and rarely the
right first move.
- Fix structure first. A bid change on a campaign containing five products teaches you nothing.
- Add negatives, so you stop paying for terms you do not want and stop competing with yourself.
- Adjust bids on terms with enough data. Under roughly 10 clicks, you are reading noise.
- Then placement adjustments, on keywords you have confirmed convert.
- Then consider up-and-down bidding, on those same proven keywords.
- Revisit weekly, not daily. Daily changes prevent the data accumulating that would justify them.
Step 3 is where discipline pays. A keyword with 4 clicks and no sale has told you nothing, and
pausing it on that basis is the most common self-inflicted wound in PPC. I’ve watched accounts pause
their way into irrelevance one under-sampled keyword at a time.
A worked sequence for a new product
Putting the pieces together, because the settings only make sense against a stage.
Weeks 1 to 3, gathering evidence. Automatic campaign plus a broad manual campaign, both on dynamic
bids down only, base bids near Amazon’s suggested range, no placement adjustments. You are buying
information, not profit. Expect an uncomfortable ACoS and resist acting on it.
Weeks 4 to 6, separating winners. Harvest converting search terms into exact-match campaigns. Keep
those on down only for now. Negate the harvested terms in the discovery campaigns. Bids on the exact
campaigns set from your break-even calculation rather than from what the discovery campaign happened to
pay.
Weeks 7 to 10, tightening. On exact-match keywords with genuine data, add a modest top-of-search
placement adjustment. Watch cost per click for a week before increasing it. Lower bids on terms
converting below break-even rather than pausing them outright, which preserves the data.
Month 3 onward, optimizing. Consider up-and-down bidding on your strongest converting keywords only.
Keep brand defense on fixed bids. Review weekly.
The most common failure in that sequence is compressing it. Sellers reach month three settings in
week two, on keywords with 6 clicks of history, and then cannot tell whether the strategy or the keyword
was wrong. In our experience patience in weeks 1 to 3 determines whether the rest of it works.
Reading the numbers without fooling yourself
Three habits that prevent most bad bid decisions.
Look at clicks before conversions. A keyword with 200 impressions and 2 clicks has a click-through
problem, not a bidding problem, and raising the bid buys more impressions of the same failure.
Segment by match type before concluding anything. Broad, phrase, and exact behave differently enough that averaging them hides the answer, and in our experience the average is what most reports show by default.
Compare against your own break-even, not against a published ACoS benchmark. I’d distrust any
benchmark that does not ask what your margin is, because the same 35% ACoS is excellent on one product
and fatal on another.
The mistakes that cost most
Bidding from the suggested bid. It reflects competition, not your economics.
Reacting to small samples. Under 10 clicks is noise. Under 30 is weak evidence.
Using up-and-down bidding everywhere. It is a tool for proven keywords, not a default.
Ignoring placement adjustments entirely, which leaves the highest-converting placement to whoever
did bother.
Treating ACoS as the goal. ACoS is a ratio, and a low one on tiny volume is not a win. What you
are optimizing is contribution after advertising, and a campaign at 40% ACoS on a high-margin product
frequently beats one at 20% on a thin one.
Never lowering bids. Bids should come down about as often as they go up. In our experience they almost never do, because lowering feels like retreating rather than like managing, and I’d treat a campaign whose bids only ever rise as unmanaged.
Where bid strategy stops helping
Worth saying plainly on a page about bidding.
Bidding cannot fix a listing that does not convert. If your click-through is fine and your
conversion is poor, more expensive clicks buy the same failure at a higher price. Our guide to
Amazon Brand Analytics covers
diagnosing which stage of the funnel is losing you the sale, and I’d do that before another bid change.
Bidding cannot fix a price problem or a stock problem either, and I’d check both before adjusting anything. An out-of-stock product with active
campaigns is spending money to advertise unavailability, which happens more often than anyone admits.
If you would rather have the account managed against margin than against ACoS, that sits inside our
Amazon PPC service. We are an Amazon Ads
partner and an Amazon SPN Verified Partner.
FAQ
What are the Amazon bidding strategies?
Three: dynamic bids down only, where Amazon lowers your bid when conversion looks unlikely; dynamic bids
up and down, where it also raises the bid when conversion looks likely; and fixed bids, where your exact
bid is used every time.
Which Amazon bidding strategy should I use?
Down only for new campaigns, discovery campaigns, and tight budgets. Up and down for proven exact-match
keywords where you know the conversion rate and your margin supports a higher cost per click. Fixed for
brand defense and controlled tests.
Do I pay my full bid on Amazon ads?
Generally no. Amazon runs a second-price style auction, so you usually pay slightly more than the next
competing bid rather than your maximum. Raising a bid mainly raises your eligibility to win placements
rather than your cost per click directly.
How do I calculate the right Amazon bid?
Start from break-even rather than the suggested bid. Divide your contribution margin per unit by the
number of clicks it takes to make a sale. An $8 margin at a 10% conversion rate supports roughly $0.80
per click before returns are considered.
What are Amazon placement adjustments?
Percentage increases applied to your bid for top of search or product pages, with rest of search taking
the base bid. Top of search converts better and costs more, and adjustments compound with up-and-down
dynamic bidding, so combining aggressive settings can push cost per click well above intent.
How many clicks before I change a bid?
Under roughly 10 clicks you are reading noise, and under 30 the evidence is weak. Pausing keywords on
small samples is one of the most common self-inflicted problems in Amazon advertising.
Last updated: August 31, 2026. Amazon’s bidding strategies, placement controls, and auction mechanics
change over time and differ by ad type and marketplace. The Amazon Ads console and Seller Central carry
the current options for your account.
