Mastering Amazon Bid Strategies: Dynamic Bids, Fixed Bids, and Bid Adjustments by Placement

Most Sponsored Products accounts are bid by feel. The advertiser picks a number that looks reasonable, watches the spend, and nudges it up or down when the cost per acquisition drifts. That works until it does not, because the bid you type into Amazon is not the bid Amazon spends. Between your number and the auction sit two separate systems that can raise it, lower it, or leave it alone, and they compound. Understanding them is the difference between steering a campaign and hoping it behaves.

There are exactly two levers. The first is your bidding strategy, which decides whether Amazon is allowed to flex your bid based on how likely a click is to convert. The second is your placement adjustment, which decides how much extra you are willing to pay to show up in the spots that convert best. Neither is complicated on its own. The trap is that they interact, and an advertiser who turns both up at the same time without doing the arithmetic can end up paying several times the bid they thought they set. Let us take them one at a time, then put them back together.

Two levers, not one

The number in the bid field is a starting point, not a final price. Amazon treats it as the most you are generically willing to pay for a click, then applies your chosen rules to it in real time. If you only ever change that one number, you are using a fraction of the control the platform gives you, and you are almost certainly mis-reading why your spend and your placements look the way they do.

The two levers do different jobs. The bidding strategy is about conversion probability: should Amazon spend more when its model thinks this particular shopper is likely to buy, and pull back when they are not? The placement adjustment is about real estate: how much more is a top-of-search slot worth to you than a product-page slot, regardless of the individual shopper? Conflating the two is the root of most bidding confusion. A campaign can be "bidding aggressively" because of its strategy, because of its placement multipliers, or because of both at once, and the fix is different in each case.

The three bidding strategies and what each one does

Sponsored Products gives you three bidding strategies, and the differences between them are mechanical, not stylistic.

• Dynamic bids - down only. Amazon lowers your bid in real time when a click looks less likely to convert, and it never raises it above the number you set. Your stated bid is a hard ceiling. This is the conservative setting, because the worst case is that you pay your bid and the common case is that you pay less.
• Dynamic bids - up and down. Amazon both lowers your bid when a conversion is unlikely and raises it when a conversion is more likely. The increases are capped: up to 100% for placements at the top of the first page of results, and up to 50% for all other placements. That cap is the part advertisers miss. Choosing up and down means you have authorized Amazon to spend up to double your stated bid in the top-of-search slot.
• Fixed bids. Amazon uses your exact bid for every opportunity and does not adjust it for conversion likelihood at all. You trade the efficiency of dynamic adjustment for total predictability of the per-click price.

The practical reading is a risk ladder. Down only is the floor of risk: it can only ever save you money relative to your bid. Fixed sits in the middle: predictable, but it keeps bidding the same amount on clicks the model already suspects will not convert. Up and down is the top of the ladder: it can deliver more conversions per dollar on the clicks that matter, at the cost of spending more, sometimes much more, on the clicks the model likes. None of these is "best." Each is correct for a different campaign state.

Placement adjustments and how they stack

The second lever is placement. Amazon reports and lets you bid separately on three placements: top of search on the first page, rest of search, and product pages. You can apply a percentage increase to any of them, up to 900%, and this control works for both automatic and manual targeting and for both fixed and dynamic bidding strategies.

Here is the interaction that catches people. A placement adjustment stacks on top of a dynamic up-and-down increase. Suppose you set a 1.00 dollar bid, choose dynamic bids - up and down, and set a 50% top-of-search placement adjustment. For a high-likelihood click in the top-of-search slot, Amazon can first apply the placement adjustment and then the up-and-down increase, and the effective bid climbs well above the 1.50 dollars the placement adjustment alone implies. The two multipliers compound rather than simply adding. This is not a bug; it is the system doing exactly what you told it. But if you do not know the levers stack, your spend will look inexplicable.

That is also why the placement report matters. Because performance is broken out by the three placements, you do not have to guess where your conversions come from. You can see whether top of search is carrying the campaign, in which case a placement adjustment there is justified, or whether product pages are quietly converting cheaply, in which case the adjustment belongs there instead. Placement bidding is one of the few levers on Amazon where you can make a decision from data rather than instinct, so use the report before you touch the multiplier.

Matching the strategy to the campaign's goal

The strategy is not a personality; it is a function of where the campaign is in its life and what you are trying to do with it.

• New or unproven campaign. Start with dynamic bids - down only. You have no conversion history, so you do not want to authorize Amazon to bid you up on clicks its model is guessing about. Down only protects you while you gather data and find which keywords and ASINs actually convert. This pairs naturally with the discovery logic in our guide to automatic versus manual Amazon campaigns, where an auto campaign on down only is the cheapest way to harvest converting search terms.
• Unprofitable campaign you want to rescue. Move to, or stay on, down only and cut bids. Up and down on a campaign that is already losing money simply loses it faster on the clicks the model likes. Fix the ACoS first, then consider scaling.
• Proven, profitable campaign you want to scale. This is what up and down is for. Once you know a campaign converts, authorizing Amazon to spend more on the highest-likelihood clicks is how you capture volume you were previously capping yourself out of. Add a measured top-of-search placement adjustment on top, and increase it in steps while watching the placement report.
• Campaign where predictable cost matters more than efficiency. Fixed bids. If you are defending a branded term, running a time-boxed promotion, or simply need a per-click price you can forecast, fixed removes the variability that dynamic introduces.

The thread connecting all four is that the bid is a profit decision. Your bid ceiling should be derived from your target ACoS and your break-even, not from what competitors appear to be paying. If you are not sure how those two numbers relate, that confusion is common and worth fixing first; we untangle it in ACoS vs. TACoS. And bidding only works if the keywords underneath it are the right ones, which is upstream work covered in keyword research for Sponsored Products.

A bidding workflow you can run every two weeks

Bids are not a set-and-forget input. A simple cadence keeps them honest without turning into daily fiddling.

1. Pull the placement report. Look at where conversions and sales actually land across top of search, rest of search, and product pages. This is the ground truth that should drive every other decision below.
2. Check each campaign's ACoS against its target. Campaigns over target are candidates for bid cuts or a move to down only; campaigns comfortably under target with capped impression share are candidates for scaling.
3. Set the strategy to match the state. New and struggling campaigns on down only; proven campaigns on up and down; predictability-critical campaigns on fixed. Change the strategy deliberately, not reflexively.
4. Adjust placement multipliers from the report, not from instinct. Raise the adjustment on the placement that is converting profitably, in steps, and re-check after the data refreshes.
5. Do the stacking arithmetic before you scale. If you are running up and down and adding a top-of-search adjustment, write down the effective top-of-search bid the two multipliers can produce, and make sure you can afford it.
6. Leave changes alone long enough to read. Bids need days of data, not hours, to show their effect. Resist the urge to re-touch a campaign you adjusted yesterday.

This cadence sits inside the broader campaign hygiene we lay out in the Amazon Ads account audit checklist, and it assumes the foundational setup is already sound, which is the job of the Sponsored Products setup guide.

The mistakes that quietly burn budget

A few errors recur often enough to be worth naming directly.

• Defaulting new campaigns to up and down. It feels aggressive and growth-minded, but on a campaign with no conversion history it authorizes Amazon to bid you up on guesses. Down only is the correct starting posture.
• Forgetting that the levers stack. Turning on up and down and a large placement adjustment at the same time, then being shocked at the spend. The multipliers compound; budget for the compounded number.
• Treating fixed bids as the safe option. Fixed is predictable, but it keeps paying full price on clicks the model already expects not to convert, so it can be less efficient than down only, not more.
• Adjusting placement multipliers without the report. Raising the top-of-search adjustment because top of search "feels" important, when the placement report shows product pages are the cheap converter. Decide from data.
• Bidding to a competitor's price instead of your own break-even. The only bid ceiling that protects profit is the one derived from your margins and target ACoS.

Get the two levers and their interaction right and bidding stops being a mystery. You are no longer typing a hopeful number into a box; you are choosing, deliberately, when Amazon may spend more, where it may spend it, and how much the two decisions compound.

Frequently asked questions

What is the difference between dynamic bids - down only and up and down on Amazon?

Down only lowers your bid in real time when a conversion looks unlikely and never raises it above the number you set, so your stated bid is a hard ceiling. Up and down does the same lowering but also raises your bid when a conversion is more likely, by up to 100% for top-of-search placements and up to 50% for all other placements. Down only can only ever spend your bid or less; up and down can spend up to double it at top of search.

When should I use fixed bids on Amazon?

Use fixed bids when predictability of the per-click price matters more than squeezing out efficiency: defending a branded keyword, running a time-boxed promotion, or any situation where you need to forecast cost precisely. The trade-off is that fixed bids keep paying your full bid even on clicks Amazon's model expects not to convert, so they are not automatically the safest choice for efficiency.

How much can Amazon placement bid adjustments change my bid?

You can raise bids by up to 900% for top of search (first page), rest of search, and product pages, and the control works for both automatic and manual targeting and for fixed and dynamic strategies. Critically, a placement adjustment stacks on top of a dynamic up-and-down increase, so the effective bid at top of search can climb well above what the placement percentage alone suggests.

Which Amazon bidding strategy is best for a new campaign?

Dynamic bids - down only. A new campaign has no conversion history, so authorizing Amazon to bid you up on its model's guesses, which up and down does, is premature. Down only protects your budget while you gather data on which keywords and products actually convert, and you can move to up and down once the campaign is proven.

Do placement adjustments and dynamic bidding stack on Amazon?

Yes. A placement adjustment and a dynamic up-and-down increase compound rather than simply add. If you run up and down with a top-of-search placement adjustment, a high-likelihood click in that slot can receive both multipliers, pushing the effective bid far above either lever alone. Always calculate the compounded effective bid before scaling.

References

• Bidding strategies for Sponsored Products, Amazon Ads Support Center
• Guide to dynamic bidding - up and down with Sponsored Products, Amazon Ads
• Adjust Sponsored Products bids, Amazon Ads Support Center
• Placement report for Sponsored Products, Amazon Ads Support Center
• Bidding rules for Sponsored Products, Amazon Ads Support Center