Your Amazon bid strategy and your placement modifiers decide what you really pay for a click, and the two multiply together. Here is what dynamic bidding actually does, what top of search is worth, and how to stop paying twice for the same position.
Amazon PPC
Your Amazon bid strategy and your placement modifiers do not add together, they multiply. A £1.00 bid with a 100% top of search modifier and dynamic up and down bidding can reach an effective ceiling of £4.00 on a single click. Most sellers set both, separately, months apart, and never work out what the combination is actually costing them.
This is the settings layer of Amazon advertising: not which keywords to target, but what happens to your money once you have chosen them. It is also the layer where the biggest silent overspends live, because nothing here shows up as an error. Everything is working exactly as configured, just far more expensively than intended.
What is on this page
- What a bid actually is on Amazon
- The three bid strategies, and when each one fits
- Placements: the three doors your ad comes through
- How the two multiply, and how to control it
- Choosing an Amazon bid strategy per campaign type
- A bidding routine that survives contact with a busy week
- Six bidding mistakes worth money
What a bid actually is on Amazon
Your bid is a maximum, not a price. You are telling Amazon the most you will pay for a click, and the auction then charges you enough to beat the next competitor rather than the full amount you offered. That is why your average cost per click sits below your bid in a healthy campaign, and why a bid that has not moved in six months can still produce a rising cost per click as the competition around it changes.
Two consequences follow. First, raising a bid does not necessarily raise what you pay, it raises how often you win, so the sensible mental model is that a bid buys frequency of appearing rather than a price. Second, when your cost per click climbs without you touching anything, the cause is outside your account: competitors, seasonality or a change in your own conversion rate feeding back into the system. The auction mechanics themselves are set out in Amazon PPC explained if you want the full version.
The three bid strategies, and when each one fits
Amazon gives you three campaign level settings, and they behave very differently.
| Strategy | What Amazon does | Best for | Watch out for |
|---|---|---|---|
| Down only | Lowers your bid when a sale looks unlikely, never raises it | New campaigns, unproven products, tight budgets | You may miss high intent placements you could have won |
| Up and down | Raises up to 100% for top of search, up to 50% elsewhere, and still lowers | Proven campaigns with a reliable conversion rate | Effective cost per click can move a long way above your base bid |
| Fixed | Uses your bid exactly, no adjustment either way | Clean testing, defensive brand campaigns | You pay full bid on clicks Amazon knew would not convert |
The practical logic is simple. Down only protects you while you do not yet know whether a product converts. Fixed gives you uncontaminated data when you are trying to learn what a keyword is genuinely worth. Up and down asks Amazon to lean harder into the moments its model likes, which is only a good trade once the listing reliably turns clicks into orders.
Where sellers go wrong is treating this as a permanent account setting rather than a stage. A campaign launched on down only in March is often still on down only in November, long after it earned the right to be more aggressive. Review the setting whenever a campaign’s conversion rate stabilises.
Placements: the three doors your ad comes through
Sponsored Products can appear in three places, and they are not variations of the same thing. Top of search sits above the organic results on the first page. Rest of search covers everything further down and on later pages. Product pages puts your ad on somebody else’s listing, usually in a carousel.
You can add a percentage modifier to top of search and to product pages, and the difference in how those two behave is stark. Top of search catches a shopper at the moment of highest intent and converts accordingly. Product pages catch somebody already reading a competitor’s listing, which is a different psychological moment entirely and typically converts less well, though it can be excellent for defensive placements on your own listings and for genuine competitor targeting.
The placement report in your advertising console gives you your own version of this chart, and it is worth pulling before you set a single modifier. Categories vary enough that a rule which works well for a homewares brand can be actively wrong for consumables.
Read placement data before you set modifiers, not after. In the advertising console, open the campaign, then the placements view. If top of search converts at twice your rest of search rate, a modifier is buying something real. If it does not, you are simply paying more for the same outcome, and the honest answer is a modifier of zero.
Not sure what your placement modifiers are actually costing you?
We will pull your placement data and show you where the multiplier is working and where it is not.
How the two multiply, and how to control it
Here is the arithmetic that catches people. Your base bid is £1.00. You set a 100% top of search modifier, so for that placement the bid becomes £2.00. You also have dynamic up and down running, so Amazon can add up to another 100% on top of that for top of search. The ceiling is now £4.00 for a click you thought you had capped at a pound.
You will not pay £4.00 on most clicks, because the auction only charges you enough to win. But your ceiling has quadrupled, and on competitive keywords the ceiling is exactly where you end up. Accounts that quietly bleed money on a handful of head terms almost always have this stacking running unnoticed.
Three ways to keep control. Pick one lever at a time: either use aggressive modifiers with fixed bidding, or modest modifiers with up and down, rather than both at full strength. Set the base bid for the worst placement, so the modifier lifts you to where you want to be for the best one instead of your base bid being a compromise. And check your actual average cost per click against your base bid monthly, because if it sits consistently above, your modifiers are doing more than you think.
Nothing in a stacked bid looks broken. Every setting is doing precisely what it was told. That is what makes it expensive.
Choosing an Amazon bid strategy per campaign type
Different campaigns have different jobs, so they deserve different settings.
Automatic discovery campaigns. Down only, modest bids, no placement modifiers. The job is finding search terms cheaply, not winning premium positions. Everything you learn here gets promoted into exact match later, which is the process in the search term report guide.
Exact match campaigns on proven keywords. Up and down, with a top of search modifier sized by your own placement data. These are the keywords you have earned the right to be aggressive on.
Brand defence campaigns. Fixed bidding. You want predictable, controlled costs on your own brand terms, not an algorithm deciding to spend more when somebody searches your name.
New product launches. Down only for the first fortnight while data builds, then reassess. There is no point letting Amazon raise bids based on a conversion model that has almost nothing to learn from yet. The full sequencing is in the first 90 days of a product launch.
Competitor targeting campaigns. Fixed or down only, with a product page modifier if the data supports it. These placements are lower intent by nature and deserve tighter control.
Want your bid strategies set per campaign rather than per account?
We run this across $112M of spend. Send your reports and we will map the right setting to each campaign.
A bidding routine that survives contact with a busy week
Elaborate bid management schemes fail because nobody keeps them up. This one takes about twenty minutes a week.
Weekly. Look at keywords with at least ten new clicks since your last review. Raise bids modestly on those converting below your target ACOS with impression share to gain. Lower bids on those converting above it. Change nothing on the rest. Move in steps of 10% to 15% rather than doubling, because large jumps make cause and effect impossible to read.
Monthly. Pull the placement report and check your modifiers still reflect reality. Compare average cost per click against base bids to catch stacking. Review whether any campaign has matured enough to move from down only to up and down.
Quarterly. Re examine the whole structure. Are your top spending keywords in campaigns with the right strategy for their stage? Has anything drifted into settings that made sense a year ago and do not now?
One discipline matters more than any of the specifics: change one thing at a time and give it a fortnight. Accounts where somebody adjusts bids, modifiers and budgets in the same session are accounts where nobody can ever explain what worked.
Six bidding mistakes worth money
Stacking a 100% modifier with up and down bidding. The four times ceiling, running unnoticed on your most competitive terms.
Leaving new campaigns on up and down. You are asking Amazon to bid harder using a conversion model built on almost no data.
Setting modifiers without reading placement data. You may be paying a premium for a placement that converts no better than the free one.
Changing bids daily. Reacting to noise, and destroying your ability to attribute any result to any change.
Doubling bids to fix low impressions. Low impressions are often a relevance or listing quality problem, and bidding harder on a listing that does not convert simply buys a worse ACOS.
Ignoring product page placements entirely. They are a real part of your spend, and for many accounts a modifier of zero there is the single fastest efficiency win available.
Get a straight answer on your bids, placements and structure
Send the last 60 days of reports. We write back with the three things we would change first.
Frequently asked questions
Down only while a campaign is new or unproven, because it caps your downside while the data is thin. Fixed bids when you are testing and want clean, uncontaminated numbers. Up and down once a campaign has a proven conversion rate and you want Amazon to lean into the placements that work. Most mature accounts end up running a mixture rather than one setting everywhere.
Amazon can increase your bid by up to 100% for top of search placements and up to 50% for other placements when it predicts a sale is likely. It only raises when its model expects a conversion, and it still lowers bids when it does not, so it is not simply an expensive version of fixed bidding.
Yes, and this is where budgets disappear. The placement modifier is applied to your base bid, then the dynamic adjustment is applied on top. A 100% top of search modifier with up and down bidding can push the effective bid on a single click to roughly four times what you thought you had set.
Usually yes, because it converts far better than the other placements in most categories, but only when the listing is ready to take the traffic. Paying a premium to send high intent shoppers to a listing with weak images or thin reviews is an expensive way to prove the listing needs work.
Start near Amazon’s suggested bid rather than above it, then adjust on data rather than instinct. Opening high buys you the top position at a price you cannot sustain and produces a spike that teaches you very little about what the keyword is genuinely worth.
Weekly for most accounts, and never on fewer than about ten clicks of new data for a given keyword. Daily bid changes mean you are reacting to noise, and worse, you can never tell which change caused which result because they overlap.
Almost always yes. Product page placements behave nothing like search: the shopper is already looking at a competitor’s listing, so conversion is typically lower and intent is different. Many accounts should run a lower or zero modifier there, though defensive placements on your own listings can be worth the money.
Your bid is a maximum, not a price. What you pay is set by the auction around you, so competitors raising bids, a seasonal surge or a change in your conversion rate can all move your actual cost per click while your setting sits untouched.
Send the right reports. Get a straight answer back.
We read the account properly and write back with where the money is going, what it is buying, and the three things we would change first. You keep it either way.
No obligation · no card · a person replies, usually the same working day