Amazon ads that stop paying for your own brand name

Mostly it is discipline and time rather than secret levers. The work that moves Amazon advertising is unglamorous and constant: harvesting converting search terms into their own campaigns, negating the ones burning budget, keeping branded and generic separated so each can be judged, and reading it all against real profitability rather than ACoS. Plenty of brands can do this in-house, and if you have somebody whose actual job it is, you may not need an agency at all. What we usually find is that it is nobody’s job, and it decays.

It is the wrong question, which is why so many accounts look healthy and lose money. ACoS compares ad spend to ad revenue and ignores referral fees, FBA costs, returns, and the organic orders the ad may have taken credit for. A 20 percent ACoS on a thin-margin product can lose money while a 45 percent ACoS on a high-margin one funds growth. The number worth managing is total profitability including fees, and the first thing we build is the view that shows it.

Sometimes, and it should be a decision rather than a default. Defending your brand term makes sense when competitors are bidding on it and taking buyers who came looking for you. It is waste when nobody is bidding against you and you are simply paying for customers who would have found you anyway. The way to tell is to test it: pause branded, watch total orders rather than ad orders, and see whether anything actually changes. Most brands have never run that test.

No, they make the problem more expensive. Amazon charges for the click whether or not the listing persuades anybody, so advertising a weak listing buys traffic for a page that cannot use it. Conversion rate also feeds Amazon’s own ranking, so a poor listing gets less organic visibility at the same time. That is why listing content is inside this engagement rather than sold separately: if the page is the constraint, the honest first move is to fix the page and spend less until it works.

They serve different buyers and should be run as one business rather than two. Amazon reaches people who search there by default and will never visit your site; your store keeps the margin, the customer relationship, and the data. The mistake is running them as separate operations with separate reports, so nobody notices when Amazon growth is quietly coming out of direct sales. We report them next to each other for that reason. Connecting the data is covered on the Amazon Ads integration page.

Structural fixes land quickly: negatives and obvious waste can be cut in the first weeks, and that alone often changes profitability without touching revenue. Rebuilt campaign architecture needs time to gather data before it can be judged, usually a month or two depending on volume. Listing changes move conversion on their own timeline. Anyone promising a specific ACoS by a specific date does not know your fee structure, your margins, or your competition, and is guessing.