Amazon PPC run against your real margin

What this actually solves

Amazon is a search engine where the results page is also the checkout, and where nobody scrolls past the first screen. A hundred sellers list the same category of product and the buyer chooses from the handful they see. Without advertising, a new product is not competing badly — it is not visible at all. That is what brings most sellers to this service.

The second problem is the one that brings them to it a second time. Amazon advertising is easy to switch on and expensive to run badly, because the money it wastes does not look like waste on the dashboard. Sales go up. Spend goes up faster. The seller sees revenue climbing and only notices at the end of the quarter that the margin has gone, absorbed by an ad bill that grew alongside the orders it created.

The number that governs all of this is ACOS — advertising cost of sale, spend divided by ad-attributed sales, expressed as a percentage. It is the inverse of ROAS, and Amazon sellers use it because they think in terms of what a sale costs them rather than what it returns. The important consequence is that break-even ACOS equals your gross margin. Take a product selling at $24.99 that leaves $8.50 once Amazon's referral and fulfilment fees, shipping and cost of goods are removed: 8.50 divided by 24.99 is 34 per cent. At 34 per cent ACOS that product breaks even. At 45 per cent it is losing money on every advertised order while the sales graph goes up and to the right.

That single calculation, done properly and per product rather than per account, is most of what separates profitable Amazon advertising from busy Amazon advertising.

What we do, and in what order

1. Access and the unit economics. Advertising-level user access granted from your account, then the arithmetic for every product worth advertising — selling price, referral and fulfilment fees, shipping, landed cost — producing a break-even ACOS per ASIN. Products whose margin cannot carry advertising are identified here and said out loud, because advertising them is a decision, not an oversight. Half a day to a day.
2. Listing and readiness review, before spend. Title, bullets, images, A+ content, review count and rating, variation structure, and whether you actually hold the Buy Box - because if you do not, your ad sends the shopper to a page that sells for someone else. Stock cover is checked against the campaign plan. Anything that will waste money is flagged before a campaign runs rather than diagnosed afterwards. Two to three days.
3. Campaign structure built for discovery and control. Automatic campaigns whose job is to find the phrases real buyers use, manual exact campaigns for terms that have already proven themselves, ASIN targeting on competitor listings where your product compares well, and Sponsored Brands for the brand term. Built in Amazon's bulk sheets so changes can be made at scale later rather than one row at a time. Two to four days.
4. The weekly cycle, run on fixed thresholds. The search term report is pulled and worked through with rules rather than intuition — terms that have spent more than about $15 with no sale become negative exact; terms running above break-even ACOS with more than about $20 spent get their bids cut or turned off; terms under roughly 20 per cent ACOS with three or more sales get harvested into their own exact-match campaign where they can be bid properly. Bids adjusted, stock checked, ASINs paused where inventory is short. Weekly, ongoing.
5. Monthly reporting against the break-even line. Spend and ACOS by ASIN set against each product's own break-even figure rather than a single account-wide target, what was cut, what was harvested, and what changes next month. Monthly, ongoing.

Tools used: the Amazon Ads console inside Seller Central, Helium 10 or Jungle Scout for keyword and competitor research, and Amazon's own bulk operation sheets for anything that touches more than a handful of rows.

What moves the price

How many ASINs. Twelve products is a job you can hold in your head. Four hundred, with variations, is a bulk-sheet operation with its own reporting structure, and the weekly cycle takes correspondingly longer.

How many marketplaces. US, UK and Germany are separate accounts with separate keywords, separate languages, separate competitors and separate fee structures. Each one recalculates the break-even maths from scratch.

Whether listing work is in scope. Ads and listings are one system on Amazon in a way they are not on Google. If titles, images and A+ content need work, that can be included - and it will show in both the price and the results.

What usually goes wrong

Ads are switched on over a listing that cannot convert. This is the defining Amazon mistake and it has no equivalent on your own website. On Google you buy a click and send it to a page you control. On Amazon you buy a click and send it to a page where your competitors' Sponsored ads sit directly beneath your product, next to their review counts and their prices. A weak listing does not merely waste the click — it hands a warmed-up buyer to the competitor advertising on the same screen. Fix what the shopper sees before paying to send more shoppers to it.

A target ACOS is picked out of the air. A number somebody heard was good becomes the target for the whole account. On a product whose break-even sits above it, that target leaves profitable sales unbought; on a product whose margin sits below it, the same target quietly loses money on every order. Break-even is arithmetic, it differs by product, and until it is calculated per ASIN there is no way to know whether a campaign is winning.

Inventory is treated as somebody else's department. A campaign keeps spending on an ASIN that has run out, which burns budget on impressions that cannot convert and gives up the Buy Box position the advertising was paying to establish. Regaining rank after a stockout costs more than the ads that were wasted during it. Stock is an advertising input on Amazon, and it belongs in the weekly checklist alongside bids and search terms.