Amazon gives you three bidding strategies and does a poor job explaining the difference between them. Most sellers pick one early and never revisit it, which means they are either leaving money on the table or paying more than they should for placements they are already winning.
This is a plain-English breakdown of what each strategy actually does, when to use each one, and how rule-based bidding changes the decision for accounts that have moved past the launch phase.
The three strategies
Amazon's bidding strategies control what happens to your bid between the time you set it and the time it competes in an auction.
| Strategy | What Amazon does with your bid |
|---|---|
| Dynamic bids: down only | Lowers your bid when conversion likelihood is low. Never raises it above your set amount. |
| Dynamic bids: up and down | Raises your bid up to 100% when conversion likelihood is high. Also lowers it when low. |
| Fixed bids | Uses your exact bid. Amazon does not adjust it in any direction. |
The key phrase in all three is "conversion likelihood." Amazon's algorithm estimates the probability that a given impression will result in a sale and adjusts accordingly. The bid you set is the starting point. The strategy determines how far Amazon can move from there.
Dynamic bids: down only
This is the default setting for new campaigns and a reasonable place to start for most accounts. Amazon can lower your bid when it predicts a click is unlikely to convert, but it cannot raise your bid above what you set.
The practical effect: you pay less for impressions that Amazon's algorithm flags as low-intent, and you pay your stated bid (or less) for everything else. There is no upside surprise on your CPC.
When it works well: campaigns with established conversion history, categories with consistent buyer intent, accounts where you have a settled sense of what your bids should be and want downside protection without giving Amazon authority to spend more.
When it falls short: the algorithm can be too conservative in new campaigns, lowering bids on placements that would have converted if you had won them. It is calibrated on historical data that does not yet exist for new products.
Dynamic bids: up and down
This strategy lets Amazon raise your bid up to 100% above your set amount when it predicts a high conversion likelihood. It also lowers bids when conversion likelihood is low.
The practical effect: your effective CPC can be significantly higher than your stated bid. If you set a $1.50 bid with this strategy, Amazon can bid up to $3.00 on your behalf when it sees a high-value placement.
Where sellers get into trouble: they look at their stated bid and do not realize their actual average CPC is materially higher. The strategy prioritizes visibility over cost control. That trade-off is worth it in specific situations, but it is not the right default for an account trying to hold ACoS.
When it makes sense: new product launches where you need impressions before conversion history exists. Time-sensitive inventory situations where winning the placement matters more than the cost. Categories where first-page placement creates a compounding rank advantage that justifies paying for it early.
When it does not: established accounts with ACoS targets, campaigns already hitting budget, or any situation where you need predictable spend. The ACoS targets guide is a useful reference before deciding how much bid flexibility you can afford.
Fixed bids
Fixed bids do exactly what the name says. Amazon uses your bid. No algorithm adjustments, no conversion-likelihood modifiers. Your stated bid competes as-is in every auction.
The practical effect: fully predictable spend. You know exactly what you are willing to pay per click, and that is what you pay when you win.
When it works well: Sponsored Brands and Sponsored Display campaigns, where the placement value is harder to assess dynamically. Manual bidding experiments where you want clean data. Campaigns being used to test keyword performance, where you need controlled variables.
When it falls short: you are leaving potential wins on the table in high-intent auctions where a small bid increase would have won the placement. Amazon's real-time conversion signals are ignored entirely, which is wasteful in categories with significant conversion variance by placement.
The case for rule-based bidding
Here is what all three Amazon-native strategies have in common: they react to Amazon's signals, not your signals. The algorithm decides when to raise or lower bids based on its estimate of conversion likelihood, not on your ACoS targets, your budget constraints, or your knowledge of when your products actually sell.
Rule-based bidding inverts this. Instead of letting the algorithm decide, you define the conditions and the response. A rule might say: if yesterday's ACoS on this campaign exceeded my target by more than 20%, reduce the bid by 15%. Or: if this campaign is hitting 95% budget utilization while ACoS is healthy, increase the budget so it does not cap out.
The difference matters for a few reasons:
You are using your data, not Amazon's. Amazon's conversion likelihood model is trained on aggregate behavior across all advertisers. Your account has specific patterns, categories with unusual buyer behavior, and seasonal dynamics that the general model does not know about. Your rules can incorporate that knowledge. The automation rules guide covers the specific conditions worth building.
You control the reaction speed. Performance rules that evaluate on prior-day data avoid the noise of real-time adjustments. Amazon's dynamic bidding reacts to each impression in real time, which creates volatility. Rules that evaluate on full-day data make steadier adjustments on cleaner signals. This matters because ACoS has natural daily variance, and reacting to every hourly fluctuation creates instability.
You can combine strategies. The most effective setups use a modest bid strategy (typically fixed or down-only) as the baseline, and layer rule-based adjustments on top. The underlying strategy provides the floor; the rules make the tactical moves.
How to choose
One mistake that cuts across all three strategies
Setting a bid strategy and never changing it. Bidding is not a one-time decision. The right strategy for a campaign at launch is often wrong for that same campaign six months later. New products that needed dynamic up/down to build history can transition to fixed or down-only once that history exists. Campaigns running rules should have those rules reviewed as the account's ACoS targets shift.
Bid strategy is part of campaign hygiene, not a one-time setup. The pre-automation audit checklist includes a bid strategy review as one of its 10 points for exactly this reason.
Frequently asked questions
What is dynamic bidding on Amazon? Dynamic bidding lets Amazon automatically raise or lower your bid in real time based on the likelihood of a conversion. Dynamic bids up/down adjusts in both directions. Dynamic bids down-only only lowers bids, never raises them. Both strategies adjust bids relative to your set bid amount, not to a target ACoS or target spend.
What is the difference between dynamic bidding up/down and fixed bids on Amazon? Fixed bids lock your bid at the exact amount you set. Amazon cannot adjust it. Dynamic up/down lets Amazon raise your bid up to 100% and lower it based on conversion signals. Fixed bids give you full control and predictable spend. Dynamic up/down gives Amazon more flexibility but can push your effective CPCs well above your set bid.
Which Amazon bid strategy is best for established accounts? For accounts with meaningful performance history, rule-based bidding gives you the most control. You define the conditions, and bids or budgets adjust on your terms rather than Amazon's algorithm. Fixed bids are a solid baseline while you gather data. Dynamic up/down is best reserved for new campaigns or product launches where you need visibility before conversion history exists.
When should I use Amazon's dynamic bidding? Dynamic bids down-only is a reasonable default for most campaigns: it prevents overpaying when conversion likelihood is low while keeping bids stable otherwise. Dynamic up/down is useful for new products or launches where you need to win impressions before conversion data exists. Once you have performance history, rule-based bidding on your own conditions gives you more predictable control.
Off Hours performance rules let you define exactly when bids and budgets should move, based on your ACoS targets and spend patterns, not Amazon's algorithm. Start a free 14-day trial.