The listing has to be able to convert first.
Advertising does not create demand. It buys visits to a page that either converts them or does not. Spend before the page is ready and you are paying to show people something that cannot persuade them.
· Advertising
There is a version of the launch story that plays out constantly. A new product goes live, campaigns go up the same afternoon, budget runs for three weeks, and the conclusion is that Amazon advertising is expensive and does not work for this category.
What actually happened is simpler. The advertising worked exactly as designed — it bought visits. The page those visits landed on could not convert them, and no bid, structure or match type addresses that.
Advertising is a multiplier, not an engine
This is the mental model worth holding. Advertising does not create sales; it multiplies whatever your listing already does with the traffic it gets. Multiply a good conversion rate and you get more sales. Multiply a poor one and you get a larger number of people who did not buy.
The arithmetic is unforgiving in a specific way. At a 10% conversion rate, a hundred clicks produce ten orders. At 3%, the same hundred clicks — the same money — produce three. You did not overpay for clicks. You paid the going rate for traffic and sent it somewhere that could not use it.
10% → ~836 orders · 3% → ~251 orders
At a ₹22 average cost per click. Identical budget, identical bidding, three times the result — decided entirely by the page, not the campaign.
What has to be true before you spend
None of this is exotic. It is the set of things that are obvious in hindsight and easy to skip when you are keen to launch.
| Element | The bar | Why it matters for advertising |
|---|---|---|
| Images | Seven or more, including lifestyle and an infographic | The main image decides the click; the rest decide the sale |
| Main image on mobile | Legible as a thumbnail | Most traffic is mobile, and a busy image is unreadable at that size |
| Title | Primary keyword present and readable | Relevance affects both what you rank for and what converts |
| Bullets | All five, written for a person | The most-read block after images |
| A+ content | Live | Meaningfully lifts conversion on considered purchases |
| Reviews | At least 10–15 before aggressive spend | Below that, paid traffic converts poorly regardless of the ad |
| Price | Within about 15% of the top competitors | Outside that band, advertising is arguing against arithmetic |
| Backend search terms | Populated, no duplicates | Free relevance you are otherwise discarding |
| Inventory | 60–90 days of cover | Running out mid-campaign loses rank you paid to build |
| Conversion rate | Around 7% or better, if relaunching | The single best predictor of whether spend will work |
For a relaunch, that last row does most of the work. You already have the number — the Business Report's unit session percentage is Amazon's own measure of it. If it is well below what your category manages, that is the project, and advertising is not.
Traffic problem or conversion problem
The Business Report separates the two reasons a product is not selling, and they need opposite responses.
- Low sessions, decent conversion — people are not finding it. This is what advertising is for, and spend here is well placed.
- Good sessions, poor conversion — people find it and decline. Advertising will buy more declines. Fix the page.
- Good sessions, poor conversion, and Featured Offer share below 100% — you are losing the Buy Box. Neither ads nor images will help; that is a pricing or seller-competition problem.
That third case is worth checking before the second, because the symptoms are identical and the fix is completely different. A week spent rewriting bullet points for a listing that was losing the Buy Box is a week gone.
Know what you can afford before you bid
The other pre-launch number is financial rather than editorial: the advertising cost of sale at which an extra order earns you nothing. That is your contribution margin — selling price less landed cost, marketplace fees and shipping.
It is worth working out before launch rather than after, because it decides whether the plan is viable at all. If your break-even ACOS is 22% and your category's cost per click implies 40% on a realistic conversion rate, the product cannot be advertised profitably at that price. Better to know that before three weeks of spend than after.
A launch can and often should exceed break-even deliberately — you are buying rank and reviews rather than profit. The distinction that matters is whether it was a decision with an end date or something you discovered later.
Start small, then scale
Two habits save real money in the first fortnight, which is the most expensive learning period any account has.
First, start at a modest daily budget rather than the number you eventually want. You are buying information about where Amazon sends your money, and that information costs the same whether you spend a little or a lot to get it.
Second, go live with a negative keyword list already in place, built from your own research — competitor brands you do not want, product types you are not, qualifiers that signal the wrong intent. The first two weeks of an automatic campaign are its most exploratory and its least discriminating, and pre-empting the obvious misses cuts that cost noticeably.
And do not cut over all at once
If you are restructuring an existing account rather than launching new, run the new campaigns alongside the old ones for the first couple of weeks rather than switching everything in one afternoon.
Cut over entirely and you lose the ability to tell whether the new structure is working or whether you simply broke something that was fine. Migrate the budget as the new campaigns show they are spending the way you expected — slower, and the only version where the result means anything.
Find out whether it is traffic or conversion.
The Business Report module reads your export and separates the two — sessions, unit session percentage and Featured Offer share per ASIN — so you know whether to buy traffic or fix the page. Free monthly allowance, and the file parses in your browser.
Frequently asked questions.
How many reviews do I need before running Amazon ads?
Somewhere around ten to fifteen before putting meaningful budget behind a listing. Below that, paid traffic converts poorly almost regardless of how good the ad or the product is, because shoppers use review count as a proxy for risk. You can advertise earlier at a small budget to gather search-term data, but treat that as research rather than as a growth attempt.
What conversion rate do I need before advertising?
For a relaunch, around 7% is a reasonable floor, though the honest answer depends on your category — some convert at 25% and some at 4%. The Business Report's unit session percentage gives you the number directly. If it is well below what comparable listings manage, advertising will multiply that weakness rather than compensate for it.
Should I fix my listing or my campaigns first?
The listing, unless the problem is genuinely that nobody is finding it. The Business Report separates the two: low sessions with decent conversion is a discovery problem that advertising solves, while good sessions with poor conversion is a page problem that advertising makes more expensive. Check Featured Offer share too — losing the Buy Box produces the same symptoms as a weak listing and needs a completely different fix.
What budget should I start a new Amazon campaign on?
Less than you intend to end up spending. The first fortnight is buying information about where Amazon sends your money, and that information costs the same whether you spend heavily or modestly to acquire it. Start small, confirm the spend is going somewhere sensible, then scale. Going in at full budget means paying full price for the least discriminating period an account ever has.
Do pre-launch negative keywords actually help?
Materially, yes. The first two weeks of an automatic campaign are its most exploratory, which means the widest and least relevant matching it will ever do. Loading obvious exclusions before launch — competitor brands, adjacent product types you are not, qualifiers signalling the wrong intent — removes a chunk of that cost before you pay for it rather than after.
How do I know if a product can be advertised profitably at all?
Compare your break-even ACOS against what your category's cost per click implies at a realistic conversion rate. If break-even is 22% and the arithmetic points at 40%, the product cannot be advertised profitably at that price — the answer is a price change, a cost change, or not advertising it. Working that out before launch is considerably cheaper than discovering it after three weeks of spend.
- The Amazon Business Report
Where sessions, conversion and Featured Offer share come from — the three numbers that decide whether a listing is ready.
- One account, eight jobs
Once the listing is ready, this is how to structure what you build on top of it.
- The weekly Amazon PPC routine
What to do every week after launch, with the thresholds that make most decisions automatic.