The Amazon Seller Central Metrics That Predict Whether You Keep Cash

The Amazon Seller Central metrics that actually predict cash retention are almost never the ones on the homepage dashboard. Most CPG operators are staring at a screen full of green checkmarks while margin quietly leaks out the side, and the account looks healthy the whole time it’s happening.

So this is the list we actually encourage operators to watch. Seven numbers, plus one that surprises almost everyone. What’s striking is what didn’t make the cut. The popular figures, revenue, ACOS, Buy Box percentage, are the ones most owners check first and the ones that tell you the least about whether any real money is reaching your bank account. We’ll get to why at the end.

1. TACoS, before you ever look at ACOS

Total Advertising Cost of Sale (TACoS) is ad spend divided by total sales, both ad and organic. ACOS only divides ad spend by ad-attributed sales, so it tells you whether your ad dollars were efficient in isolation and nothing more. TACoS tells you the thing that actually matters for a repeat-purchase brand: whether your ads are building durable organic rank or whether you’re just renting your position forever.

We watch the direction, not a single reading. As organic compounds, TACoS should drift down over quarters, and that downward trend is the flywheel doing its job. When TACoS sits flat or climbs while revenue stays flat, you’re paying rent, and you’ll keep paying it every month you don’t fix the underlying rank.

2. Break-even ACOS

This one equals your product’s margin percentage, and it functions as a ceiling. Above it, every ad sale loses money. Simple as that.

Here’s where owners get quietly hurt. Most set their bids off what competitors appear to be bidding rather than off their own break-even number. That means a competitor with a fatter margin can profitably outbid you into a loss, and nothing on the dashboard flags it. We’ve seen a brand hold an aggressive bid for months to “stay competitive” on a hero keyword, never realizing every sale on it was underwater, because the campaign looked productive and the sales kept coming. It stayed busy the entire time it was losing money, and the account had no reason to flag the gap.

3. Landed margin per unit, after the full fee stack

Gross margin off your COGS spreadsheet is not landed margin. The real number stacks the referral fee, which runs 8 to 15 percent depending on category, plus the FBA fulfillment fee set by your product’s size and weight tier, on top of COGS.

That size and weight tier is where launches go wrong. Get the box dimensions wrong at setup, cross a dimensional threshold you didn’t know existed, and you can sit underwater on a SKU that reads as profitable in every internal model for a year before anyone reconciles it against the fee reports. We learned to check this at launch the expensive way, after watching a “winning” product turn out to have been losing a little on every unit the whole time. The fee was baked in on day one. Nobody looked until the annual margin review forced it.

4. Returns rate, as a planned line and not a surprise

Returns belong in your unit economics from day one, sized to your category. Anything with fit, sizing, or fragile shipping runs higher and needs a bigger allowance. Build the return processing fee into your model deliberately.

Left out of the model, they resurface as an unexplained margin miss every quarter instead of a cost you forecasted and planned around. Same dollars either way. The difference is whether you saw them coming.

5. Storage cost as a share of revenue, plus long-term storage fees

Track storage as a percentage of revenue and watch it for creep. When that share starts climbing, the cause is almost always aged inventory that has crossed into the long-term storage window and started compounding fees against you.

Most owners watch total inventory value instead of inventory age, so the long-term fee hit lands as a shock. For a brand carrying several flavors, formulas, or shades, that aging tends to concentrate in one or two slow movers, quietly running up the bill while the top sellers keep the aggregate looking fine.

6. IPI, the Inventory Performance Index

This is the number that gets ignored right up until it becomes a crisis. IPI is Amazon’s inventory-health score, and a low one triggers restock limits that cap how much inventory you’re allowed to send in.

Picture the timing. You’re scaling ad spend into your best-ranking SKU, demand is climbing, and a restock cap stops you from sending enough units to keep up. You stock out at exactly the wrong moment. The velocity you lose resets the organic rank you spent months and real ad dollars earning, and clawing it back costs you both again. Owners watch “units in stock,” not IPI, so the cap blindsides them mid growth push, during the one stretch where running out hurts most.

7. ODR and overall Account Health Rating

Order Defect Rate wants to sit under 1 percent, and Account Health tracks it alongside policy compliance and shipping performance. This isn’t a growth metric. It’s existence insurance. A suspended account converts nothing and ranks nowhere, no matter how strong your listing or how efficient your ads.

It’s also the dashboard most owners never open until a deactivation email forces them to. A slipping Account Health rating is the earliest warning you get that the whole operation could go dark, and it’s worth a genuine weekly glance even though it says nothing about this month’s margin.

The bonus metric: fee leakage nobody reconciles

Here’s the metric that surprises people. A meaningful slice of cash sits in Amazon’s favor by default, and it never appears as a line on your P&L unless someone goes looking for it. Fee miscalculations. Missed reimbursements for FBA inventory that was lost or damaged in Amazon’s own network. Returns that came back mislabeled or never came back at all.

None of it shows up on the top-line reports. It’s real money, it’s yours, and recovering it takes someone deliberately pulling the reports and filing the claims. Left alone, it simply stays with Amazon, quarter after quarter.

Why most owners watch the wrong Seller Central numbers

The instinct is to watch what’s easiest to see in Seller Central. Revenue, ACOS, Buy Box percentage all sit right on top of the account, updating in real time, asking nothing of you. So they’re what gets watched.

They’re also surface readings. Revenue with no margin behind it is vanity. ACOS with no TACoS beside it hides whether you’re buying durable rank or just buying today’s sale. The numbers that actually predict whether you keep cash, the TACoS trend, landed margin, IPI, ODR, and that unreconciled fee leakage, all sit one report deeper or need a little math to assemble. None of them greet you on the homepage.

So they get skipped, right up until they cause a problem visible enough that you can’t skip them anymore. By then it’s a recovery project, a suspension appeal, a stock-out you’re digging out of, a year of thin margin to explain. The whole point of watching the right numbers is to keep them as monitoring lines instead of emergencies. Green checkmarks are not the same as kept cash, and only one of them pays for the next production run.