Quick answer: When Amazon ads spend without sales, first check reporting lag, stock availability, Featured Offer ownership, listing status, and price, then review search-term relevance, click volume, conversion, placement, bids, and margin. Don't judge a recent campaign before attributed sales have had time to populate.
A campaign spending money without sales feels simple: the ads are bad. Sometimes they are. Sometimes the ad did its job and the offer failed after the click. Sometimes Amazon hasn't finished attributing the sale. Sometimes another seller holds the Featured Offer. Pause the wrong thing and you can hide the symptom without fixing the business.
The 12 checks, in order
1. Is the data mature enough? Amazon Ads sales don't always appear at the same time as spend and clicks. If the clicks happened today, the first action is often to wait, not optimize, covered further in how often does Amazon Seller Central update.
2. Is the advertised product in stock? An ad can keep serving while the offer has limited stock or a weak delivery promise. Check available FBA quantity, reserved and transfer quantity, and the delivery date shown to a shopper.
3. Do you hold the Featured Offer? On a shared ASIN, traffic can land on the detail page while another seller wins the purchase. If Featured Offer percentage fell, advertising may be paying to send shoppers to a page where a competitor receives the order, covered in Amazon Buy Box percentage dropped.
4. Is the price competitive at the moment of the click? A campaign report shows a period; the shopper saw one moment. Price may have increased, a coupon may have expired, or a competitor may have discounted since the report was pulled.
5. Is the search term relevant to the exact product? Broad and automatic targeting can find terms that look related but carry the wrong intent. A "glass jar" ad may receive clicks from shoppers looking for a different volume or a single unit rather than a case of 24. Use the search term report, not only the target list.
6. Do you have enough clicks to judge conversion? Zero sales after three clicks tells you almost nothing. Zero sales after 80 relevant clicks is different. Expected orders equals clicks times expected click-to-order conversion rate: at 10% conversion, 20 clicks produces an expectation of about two orders, not a guarantee. Avoid rigid rules like "pause after ten clicks" across every price and category.
7. Did detail-page conversion fall outside advertising? Check sessions and unit session percentage. If overall detail-page conversion fell, the campaign may be exposing a listing problem rather than creating it, investigated further in Amazon conversion rate by ASIN.
8. Are clicks concentrated in an expensive placement? Top-of-search traffic can convert differently from product-page or rest-of-search placements. A placement shift can raise CPC faster than conversion improves.
9. Is the bid buying the wrong auction? A high bid can push the campaign into more expensive or broader auctions. A low bid can leave it with scattered, low-volume traffic. The goal isn't the lowest CPC, it's profitable or strategically justified traffic.
10. Are sales landing on another variation? A shopper may click one child and purchase another. Review the campaign's promoted and purchased-product reporting where available, especially for color, size, or pack-count variations.
11. Is the product economically capable of supporting the campaign? Calculate pre-ad contribution per unit: selling price minus Amazon fees, product cost, inbound cost, and expected return cost. That amount sets an approximate break-even ad cost per order, a calculation covered further in TACOS vs ACOS vs ROAS.
12. Is the campaign supposed to make a direct sale? Some campaigns are launched for rank support or product discovery. That doesn't excuse unlimited loss, but it changes the evaluation. State the purpose, budget, and stopping condition before launch.
The decision table
| Pattern | Likely issue | First action |
|---|---|---|
| Spend today, no sales yet | Reporting or attribution delay | Wait for mature data, check for runaway spend |
| Clicks up, Featured Offer down | Offer ownership | Fix price, stock, delivery, or eligibility |
| Relevant clicks, conversion down everywhere | Listing or offer | Audit retail page and recent changes |
| Irrelevant search terms | Targeting | Add negatives and tighten match types |
| Good conversion, terrible ACOS | CPC or margin | Reduce bids, change placement, improve economics |
| Sales on other children | Variation behavior | Review purchased-product and child performance |
Waiting solved the wrong problem
A homeware seller saw $180 of Sponsored Products spend and no sales on a new campaign, and paused every keyword. The next day, four orders appeared in attribution. The campaign still wasn't good, but the diagnosis changed: two orders came from a highly relevant term, one from a vague automatic target, and one purchase went to a larger pack size. Restarting the relevant target, excluding the vague term, and moving budget toward the variation customers actually bought fixed the actual problem. Waiting didn't solve the campaign. It prevented solving the wrong one.
EcomSanity places ASIN sales, FBA inventory, days of stock, conversion, and Featured Offer percentage in one view, so the first question after checking ad spend doesn't require opening five separate reports. It doesn't replace search-term optimization, but it helps identify when the advertised ASIN itself has become a weak destination.
Frequently asked questions
How many clicks should an Amazon ad get before a sale?
There's no universal number. Use the product's expected conversion rate: at 10% conversion, one order per roughly ten clicks is an average expectation across enough data, not a guarantee for each ten-click block.
Should I pause an Amazon campaign with no sales?
Pause or reduce it once the data is mature, the spend is economically meaningful, and the traffic has had a fair chance to convert. First rule out reporting delay, stock, Featured Offer, and listing problems.
Why do Amazon ad sales appear late?
Conversions can occur after the click and reporting can take time to populate. Attribution windows and later adjustments mean recent data is not final, sometimes for hours after the click happened.