How to Manage Amazon PPC: 5 Decisions Before Any Bid Change
How to manage Amazon PPC: five decisions to make before any bid change, from offer checks and product budget to effective bids and reading results.
Last updated 8 min read
To manage Amazon PPC well, make five decisions in order before you change a single bid: whether the problem is in the ads at all, which products deserve ad money this week, where the money is mismatched, how big the change should be, and how you will know it worked. Most Amazon PPC problems are decision problems, and the bid is where those decisions end up, not where they start. Start this week with the first decision: before you open the bid screen, confirm that your top products can still be bought at the price you set, and that the price still holds up against competing products.
The pressure to change bids usually starts with something that looks urgent: spend falls overnight, campaigns run out of budget every evening, or the ad reports look fine while total sales slip. Bids are the easiest lever in the console, so they get pulled first. A bid change can't make an unbuyable product sell, can't be judged on orders Amazon hasn't counted yet, and can cost more than the number in the bid field suggests. A workable Amazon PPC strategy puts the bid at the end.
Decision 1: Is it an ads problem at all?
Often it isn't. Many sudden drops start outside the ad account, so rule out the offer before you touch the ads. The fastest test is organic orders: if they fell at the same moment as ad sales, the problem sits on the product side. The product may have become unbuyable, lost the Featured Offer (the Buy Box, meaning the offer that gets the Add to Cart button), or run out of stock.
In one account we audited, two top-selling variations became unbuyable for about two days. The campaigns advertising them normally carried close to half of the account's ad spend. Organic orders dropped to zero at the same time, and nothing in the ad console had changed. Raising bids on a product that can't be bought does nothing; the offer has to be fixed first.
Price comes next. When a promotion ends, the checkout price steps back up and ad sales fall with it, so check the daily average selling price before concluding the ads got worse. If conversion rate falls while clicks and cost per click hold steady, compare your price with the competing products shown next to yours; a competitor's price cut or coupon makes the same clicks convert less often, and a higher bid only buys more of them.
If the offer is clean, check what changed in the account. Amazon's change history shows what changed and when, but not who made the change, so edits from your team, a bid tool and an automated rule all look the same there. For a bad week, work through what to check when Amazon ad sales drop suddenly before anyone touches a bid.
Decision 2: Which products deserve ad money this week?
A product deserves ad money this week only if it can be bought, has sellable stock, is priced to compete and still earns a margin after ad costs. Treat ad spend like working capital: it should move to wherever it can turn into profitable sales, and the list will change as stock, prices and promotions change.
Then compare what you advertise with what actually sells. In a large catalog, coverage drifts in both directions: products that sell get no ads, and advertised products stop selling. In one catalog, roughly a quarter of the products that sold in the last 90 days had no active ads, while nearly twenty advertised products hadn't sold at all.
Only then look at spend by intent, meaning how the money splits across brand searches, ads on your own product pages, generic searches and competitor targets. The layer with the best return can be the one getting the least money. In a catalog that changes every week, repeat the product check weekly and allocate ad budget across the catalog only among the products that pass.
Decision 3: Where is the money mismatched?
The mismatch usually sits in placement, time of day and budget, audience, or keywords and match types. Each has a quick test that costs less than a bid change.
| Where | What to compare | The common trap |
|---|---|---|
| Placement | Total impressions before and after a placement change | Raising the extra bid for the top search slot mostly moves impressions you already had into that slot |
| Time of day and budget | The hour campaigns run out of budget, and how those hours convert | Running out every evening is a schedule nobody chose, and the hours lost often convert worst |
| Audience | The share of impressions that match the audience | A bid boost applies only to matching impressions, so a low match rate buys almost nothing |
| Keywords and match types | The search terms behind each keyword's spend | A broad keyword with your brand name in it can spend most of its money on generic searches |
A top-of-search multiplier is the extra percentage added to your bid when the ad competes for the top of the first page of search results. In one campaign, raising it doubled the top-of-search share of impressions, from about a third to about two-thirds, while total impressions stayed flat. The extra money bought a better slot for roughly the same impressions.
When budgets run dry every evening, price the lost hours before you raise the budget or lower bids in weak hours. In keywords, look for leftovers from earlier setups, such as negative exact keywords that block the campaign's own exact keywords.
For an outside read, a free read-only audit shows where spend is wasted and how your campaign structure holds up.
Decision 4: How big should the change be?
Size the change from your goal and from the effective bid, meaning what your ad actually bids once every adjustment is applied.
The goal sets the limit. When the goal is growth, a larger step can make sense as long as each extra dollar still brings back more in sales than your break-even ROAS, which is one divided by your margin before ad costs. When the goal is efficiency, trim where extra spend returns less than your target, in steps you can read before taking the next one. Cuts have a floor: once a bid falls below the price that wins a placement, a keyword can go from steady volume to zero impressions within days.
Then work out what the bid really is. Placement adjustments and schedule-based bid rules (rules that raise bids during set times) add together rather than multiply: a $1.00 bid with a +10% placement adjustment and a +20% rule is $1.30, not $1.32. Dynamic bidding is another layer: with dynamic bids (up and down), Amazon's guide to dynamic bidding says Amazon can increase or decrease those bids by up to 100% for all placements, based on performance. Check how your tool computes the effective bid before you size a step.
The goal also moves with price. A discount raises the ROAS you need to break even, which is why deal math for Prime Day starts from margin before any bid goes up.
Decision 5: How will you know it worked?
Decide how you will read the result before you make the change: compare the changed targets with targets you left alone, wait for the data to settle, and use total sales and organic rank as guardrails.
Judge the extra dollars on their own with marginal ROAS: the sales the additional spend brought in, divided by that additional spend. Blended ROAS, the average across all spend, can look flat while the extra dollars return far less. In one account we manage, a round of bid increases left the changed targets' ROAS looking flat. Compared with untouched targets over the same days, each extra dollar returned about $1.70 in sales, against a goal of $3.
Give the data time, too. Amazon keeps crediting orders to ad clicks for days afterward, so recent days always look worse than they will. When we measured this in an account we manage, yesterday's attributed orders were around 80% of their final count, and the count took about a week to settle. Cutting bids because of the last three days punishes orders that haven't arrived yet.
Then watch the business around the ads. TACoS (ad spend divided by total sales, organic included) and organic rank (where the product shows up in unpaid search results) tell you whether ad growth is adding to the business. If ad spend grows faster than total sales, TACoS rises, and that is the moment to ask whether the ads are adding orders or claiming ones you would have had anyway.
How to manage Amazon PPC changes so they can be verified
Every change should be read back from Amazon, logged where you can read it, and stopped when someone else has already edited the same setting. When several people and tools edit the same account, some changes will land outside anyone's log, and a process that simply overwrites them hides the conflict. Automation should also refuse to run when the account no longer matches its plan.
When a client's team hand-edited a few bids that our automation managed, the system kept the client's values and raised an alert instead of putting its own values back. Ask any vendor to show you that behavior and the other guardrails Amazon PPC automation should have.
Used as an Amazon PPC management framework, the five decisions also make a monthly review: for each one, you should be able to point to where it was made last month and what the numbers showed. A decision nobody can point to is the one to fix first.
What to check this week
- Buyability by variation. In Seller Central, open Business Reports, then Detail Page Sales and Traffic by Child Item. Scan the last 14 days for any top variation whose units or Featured Offer percentage fell toward zero. A check run today can't show you yesterday.
- Enabled campaigns with no spend. In Campaign Manager, filter to enabled campaigns, set the date range to the last 30 days and sort by spend. For each campaign at zero, confirm its product can be bought before changing the bid.
- Ad coverage against sales. List the products in your active campaigns next to the products that sold in the last 90 days. Mark products that sell but have no ads, and advertised products with no sales.
- Placement changes. In the placement report, pick a campaign whose top-of-search adjustment went up recently and compare total impressions before and after, alongside top-of-search share.
- Your last bid change. Compare the changed targets with similar targets you left alone over the same dates, and leave out the most recent week.
If you want an outside read before the next round of bid changes, start with a free Amazon PPC audit.