You get a monthly report. It has a screenshot of your ad dashboard, a paragraph about "continued optimization," and a chart of total sales that goes up because your category is seasonal and August always beats June. Nothing in it tells you whether your agency did anything, or whether your brand just rode the calendar.
This is the quiet failure mode of Amazon management. Not fraud, not incompetence you can point to, just a slow drift into busywork that produces activity without producing decisions. The brand owner keeps paying because the account isn't broken. But "not broken" and "actively managed" are different things, and the gap between them is where margin disappears.
Start With What the Report Doesn't Show
Most agency reports lead with top-line revenue, ACoS, and maybe TACoS. Those numbers matter, but they describe what already happened. They don't tell you why, and they don't tell you what changes next.
Ask your agency for the report they'd be embarrassed to show you: bid changes made in the last 30 days and the reasoning behind each one, search terms harvested and negated, listings edited and what specifically changed. If that report doesn't exist, or takes a week to produce, you have your answer. Real management leaves a paper trail because decisions get made every week, not once a quarter. This is the same discipline behind a good search term report review: harvest, negate, promote, watch, on a schedule, not when someone remembers.
The Metrics That Separate Signal From Noise
Revenue and ACoS are lagging indicators. They tell you the score at the end of the game. What you want are the numbers that predict the next 30 days, not summarize the last 30.
- Contribution margin per ASIN, not blended account revenue. A product can grow sales and lose money at the same time, and revenue alone will hide it. If your agency can't produce this number by product, they're not managing profit, they're managing top-line vanity. This is the whole argument behind why contribution margin should drive every Amazon decision, and it's the single fastest way to tell a real operator from a spend manager.
- Click share and conversion rate by keyword, not just overall CVR. Overall conversion can hold steady while your best keywords quietly lose share to a competitor. You won't see it unless someone is watching keyword-level data every week.
- New-to-brand and repeat purchase behavior, if you sell a repeat-use product. An agency optimizing only for immediate ROAS will happily overspend to acquire customers who were never going to come back, and underspend on the ones who would.
- Inventory health against sell-through, because a stockout erases weeks of ranking work in days. If reporting never mentions days of cover, nobody is watching the thing most likely to tank your account overnight.
If your monthly report could have been written without anyone logging into Seller Central that week, it isn't a report. It's a summary.
Ask for the Target ACoS Logic, Not Just the Number
Any agency can tell you your ACoS is 22%. Few can tell you why 22% is the right number for that specific product, at that specific stage, given its margin. A defensible target ACoS comes from a formula built on contribution margin, adjusted for whether the product is launching, growing, or mature. If the number in your report hasn't moved in six months despite the product moving through different lifecycle stages, nobody recalculated it, they just copied it forward. The full method for building this correctly is laid out in the right way to set a target ACoS for each product, and it's worth comparing against whatever your current number is based on.
Watch for Siloed Work
A common pattern with underperforming management: PPC, listings, and creative are handled by different people (or different vendors) who never talk to each other. Bids get adjusted on a listing whose main image was never tested. A new A+ module goes live the same week PPC budget gets cut, so nobody can tell which change moved conversion. Each piece looks fine in isolation. Together, they don't compound, because nothing is coordinated.
This is worth checking directly: ask what changed on the account in the last two weeks across advertising, listings, and inventory, and ask whether those changes were sequenced on purpose. If the answer is vague, the account is being run as four disconnected projects instead of one coordinated system, and you're paying for four sets of effort that partially cancel each other out.
The Questions That Expose Real Management
You don't need to audit spreadsheets to find out if you're getting real management. Three questions, asked directly, usually surface the truth fast:
- "What did you change this week, and why?" A real manager has a specific answer involving a specific ASIN or campaign. A vague answer about "ongoing optimization" means nothing happened, or nobody's tracking what did.
- "What's the target ACoS on my top three products, and how was it calculated?" If the number can't be traced back to margin, it's a guess dressed up as a strategy.
- "What's the biggest risk to this account right now?" A real operator has an answer, usually inventory, a compliance flag, a competitor move, or a keyword losing share. If they say "nothing, everything's good," they aren't watching closely enough to know.
None of these questions require Amazon expertise on your part. They just require you to expect a specific answer instead of a comfortable one.
What to Do This Week
Pull your last three monthly reports and check them against one standard: does each report contain a decision, or just a description? A decision looks like "we cut bids on Campaign X by 15% because ACoS exceeded target for three consecutive weeks." A description looks like "sales trended upward, we continued to monitor performance."
If most of what you're reading is description, ask for the three questions above in your next call, and ask for them in writing. The answers will tell you, faster than any dashboard, whether your account is being managed or just being watched.