Every Amazon account has things that underperform: a keyword that spends without selling, a campaign whose ACoS has crept up, an overnight stretch that buys clicks no one converts. The two standard responses are to pause it or to lower the bid. Both cut spend. They do not cost the same.
This post compares the two: what each does in the auction, what each puts at risk, and a simple way to choose between them using your own numbers.
What pausing does
Pausing takes the target out of every auction. Spend goes to zero, and so do impressions, clicks and sales. Nothing is learned while it is paused, and nothing is wasted.
Pausing is clean and easy to reason about. Its cost is the sales you give up. If the paused target produced a few orders at a poor ACoS, those orders are gone. If it produced none, pausing costs nothing.
What lowering the bid does
A lower bid keeps the target in the auction but wins fewer of them, usually in less visible placements. Clicks fall, cost per click falls, and the clicks you still get tend to come from placements and shoppers where your ad competed at a lower price.
Lowering the bid keeps some sales at a better cost. Its risk is that it can also keep some waste. A target that does not convert at any price will not start converting at a lower bid; it will just waste money more slowly. How Amazon PPC bidding works covers how bids translate into placements.
The deciding question: does it convert at any price?
The choice comes down to one question. Does the target produce orders, just too expensively? Or does it produce almost none?
If it produces orders at a poor ACoS, lowering the bid is usually better. The orders are real; you are paying too much for them. A lower bid brings the cost back toward your target and keeps the sales.
If it produces few or no orders after meaningful clicks, pausing is usually better. There is no cost at which zero orders becomes a good ACoS. The same applies to search terms that do not fit the product, which belong in negative keywords rather than a lower bid.
Applying it to hours of the day
The same logic decides overnight scheduling. Pull four weeks of hourly data and compare each hour's conversion rate to the daily average.
Hours that convert far below average, for example under half, with enough clicks to trust the number, are candidates to pause. They are spending without selling.
Hours that convert somewhat below average are candidates for a lower bid. They sell, just at a worse rate. Cutting the bid in those hours keeps some of the orders at a cost closer to your target.
Illustrative example: Harbor Kitchen sees that its hours from around midnight to 5am convert at a fraction of its daily rate, while 6am to 8am convert somewhat below average. It pauses the first block and lowers bids in the second. The overnight strategy guide covers this pattern in more detail.
Applying it to campaigns and keywords
Whole campaigns. Pausing a campaign is a large move. Before doing it, check whether a few targets inside it are causing the problem. Often a campaign with a poor ACoS contains a handful of good keywords and a few expensive ones. Fix the expensive ones and keep the campaign.
Keywords and product targets. These are the easiest to decide individually. Converting but too expensive: lower the bid. Not converting after enough clicks: pause it or add it as a negative.
New targets with little data. Do neither yet. A target with a handful of clicks and no orders has not had a fair test. Set a click threshold, based on your conversion rate, before you judge.
What each one risks
Pausing risks lost sales and lost learning. A paused keyword gathers no new data, so you will not see if it starts converting later. Long pauses on whole campaigns can also mean a slower restart.
Lowering bids risks slow waste. A target kept alive at a low bid can spend small amounts for months without producing anything. It also risks losing placements that mattered: if the target's conversions came mostly from top of search, a lower bid may lose exactly those.
Neither risk is large if you review the change. Both become large if you make the change and forget it.
Reversibility
Both are reversible, which is why neither decision needs to be perfect. Pausing is the easier one to undo cleanly: turn it back on. A lowered bid is easy to raise but harder to evaluate, because performance at the new bid mixes with everything else that changed.
Write down what you changed and why. A simple log, even a spreadsheet, turns every pause and bid cut into something you can review. Keeping a change log explains what to record.
A simple rule of thumb
Lower bids when the target sells too expensively. Pause when it does not sell. Wait when you do not know yet. Review either change after one to two weeks of data, and undo it if the result is worse than what you had.
Frequently asked questions
Does pausing an Amazon campaign hurt its performance later?
Short pauses, such as overnight, do not appear to cause lasting harm in most accounts. Long pauses mean the campaign gathers no new data, and when it restarts it may take time to return to its previous delivery. Watch the first days after a long pause before judging the campaign.
Is it better to lower bids overnight instead of pausing?
It depends on how weak the overnight hours are. If they convert far below the daily average, pausing saves the most. If they convert somewhat below average, a lower bid keeps some sales at a better cost. Many accounts use both: pause the deadest hours and lower bids in the weaker ones around them.
How much should I lower a bid instead of pausing?
Lower it in proportion to how far the target's conversion rate falls below what you need. A keyword converting at half the rate you need for your ACoS target can usually support roughly half the bid. Make the change in steps and review after a week or two of data.
Off Hours dayparting rules can pause campaigns in dead hours on a 15-minute cadence and log every change, so you can see exactly what each pause did. Start a free 14-day trial.