Amazon's Average TACoS, Calculated From Amazon's Earnings Reports

If your TACoS rose this year, the first question in your next review will be whether the rise is yours or the market's. Amazon's own earnings reports can answer that, and they point one way: across the whole marketplace, ad spend is taking a bigger share of every sale, year after year.

Every other TACoS benchmark you have seen comes from a vendor survey, an agency's client base or a forum thread. None are audited, most don't disclose their sample, and no two agree. Amazon's quarterly filings cover every seller and every sale. They don't state a TACoS, but they contain enough to estimate one, and the direction it shows is worth planning around.

Quick Answer

Advertising is taking a steadily bigger share of what shoppers spend on Amazon. Measured across the whole marketplace, ad spend was a little under 10% of sales in Q3 2026, and that share has grown by roughly half a point to almost a full point every year. Some of the increase is Amazon selling new kinds of ads, like Prime Video and live sports, rather than sellers paying more for the same clicks. But the direction is clear: holding TACoS flat now takes work, and a plan that assumes it will stay flat on its own is likely to miss.

What Amazon's Numbers Mean for Your TACoS

Three things follow from the trend, and they hold however you adjust the details.

Some of your TACoS drift is the market's doing

If your TACoS rose by about the same amount as the market's over the same year, most of that rise came from a more crowded auction, not from a mistake in your account. That changes the conversation in a quarterly review from "who broke efficiency" to "do we accept a more expensive auction, or reallocate."

Holding TACoS flat now counts as outperforming

With the market drifting up every year, holding your TACoS steady means you outperformed the market. A brand whose TACoS rose half a point while the market rose more than that did better than the headline suggests. Set goals against the market's movement, not against last year's number.

Your own number can sit well above the average

A market-wide figure near 10% blends every category, price point and stage on Amazon. TACoS in the mid to high teens is common right now, particularly for launches, competitive categories and brands buying growth. Use the market number to judge how fast you are drifting, not where you should land.

How Sellers Can Use the Market Trend

Compare your year-over-year change to the market's

Take your TACoS for a quarter and the same quarter last year, and compare the change to the market's change in the table further down. Always compare the same quarter year over year. Prime Day moved from Q3 into Q2 this year, so comparing one quarter to the next would mislead you.

Build the drift into next year's plan

If your margin model assumes TACoS stays flat, stress-test it at half a point to a full point higher. That range brackets the market's recent yearly rise, and it shows which products still clear their margin targets if the auction keeps getting more expensive.

Decide which SKUs earn the extra cost

When the auction gets more expensive across the board, spreading spend evenly lets every product absorb the increase. Products with healthy margins and strong conversion can carry a higher TACoS. Thin-margin products may need tighter targeting, a lower share of spend, or a lean on organic rank instead. For the levers, see our guide on how to lower TACoS without killing sales.

How Much Weight to Give These Numbers

The estimate is useful for direction and less useful for precision. Three things decide how much it should move your plans.

Trust the direction more than the exact level

The method relies on one assumption, explained below, that can move the level by about a point either way. The direction and pace barely change under any reasonable version of that assumption. Quote "ad share of sales is rising by around half a point to a point a year," not "market TACoS is 9.75%."

Part of the rise isn't seller money

Amazon's advertising revenue includes Prime Video ads, live sports such as NFL and NBA, Twitch and DSP ads on other websites, alongside the sponsored ads you buy. Amazon reported that its live sports ad inventory sold out in Q2 2026, and Prime Video ads were not a meaningful business before 2024. All of that lifts the ratio without any seller paying more per click. Amazon doesn't split its ad revenue by type, so nobody outside Amazon can say exactly how much of the rise this explains.

It's quarterly and lagging

Amazon reports about a month after each quarter ends. The estimate tells you where the market has been, not what it is doing this week. Use it for quarterly reviews and annual planning, not for day-to-day bid decisions.

How the Market TACoS Is Calculated

This section explains where the numbers come from, for anyone who wants to check them or rerun them. You don't need it to use the takeaways above.

Why you can't read TACoS straight off Amazon's report

Most published TACoS benchmarks are built from a sample of accounts. Amazon's filing covers everything, but it reports sales in a way that is easy to misread, and getting this wrong produces a number that looks alarming and means nothing. Amazon's report splits its revenue into seven lines. Three matter here:

Source: Amazon Q3 2026 earnings release, supplemental tables.

The middle row is the trap. When a third-party seller makes a $30 sale, Amazon reports only the fees it kept, perhaps $10 of it. The other $20 never appears anywhere in the filing. Third-party sellers make most of the sales on Amazon (61% of units in Q3 2026), so most of the marketplace's sales are missing from the report.

That is why the obvious calculation misleads. Divide ad revenue by the two sales lines as reported and you get about 17%. Plenty of individual sellers run a TACoS that high, but as a market average it is far too high, because the bottom half of the ratio is missing most of the sales.

Filling in the missing seller sales

The fix uses one idea: if you know roughly what share of each sale Amazon keeps in fees, you can work backwards from the fees to the sales. That share is often called the take rate. Here is the whole calculation with the Q3 2026 numbers, at an assumed 35% take rate:

The one assumption, and why it matters less than it looks

Nobody outside Amazon knows the exact take rate. Once referral, fulfillment and storage fees are added together, a reasonable working range is 30% to 40% of the selling price, and your own settlement reports show your actual blended rate. Running the calculation across that range moves the answer by about a point either way:

The trend barely moves, though. The same assumption applies to every quarter, so it raises or lowers all of them together. Between Q2 2025 and Q2 2026, market TACoS rose 9.3% in relative terms at a 30% take rate, 9.3% at 35% and 9.4% at 40%. That is why the direction is the reliable part.

The trend, quarter by quarter

At a 35% take rate, across the six quarters in Amazon's current report:

Q3's smaller rise is mostly calendar. Prime Day ran in Q3 in 2025 and in Q2 for most large countries in 2026, which moved ad spend and sales between the two quarters. Combining Q2 and Q3 cancels that out and shows a rise of about 0.6 points year over year, slower than the first half of the year suggested but still clearly upward.

The longer view

On a full-year basis the same pattern goes further back: about 7.4% in 2023, 8.1% in 2024 and 9.0% in 2025, nearly identical across the take-rate range. Put another way, of every $100 in new marketplace sales, advertising captured about $15 in 2024 and about $18 in 2025.

Running it on your own account

Pull the supplemental table from Amazon's quarterly earnings release (on Amazon's investor relations site, or in the 10-Q on SEC EDGAR). Run the three steps in the box above for the same quarter this year and last, ideally using your own blended fee rate from your settlement reports. Then compare the change to your own TACoS for the same quarters and marketplaces. It takes about fifteen minutes in a spreadsheet, once a quarter.

One example: a brand whose Q2 TACoS went from 10.5% to 11.4% (up 0.9 points) saw the market rise about 0.8 points over the same year. Nearly all of its increase was the market, and the question for its review became which SKUs could absorb a more expensive auction.

Finding Where Your TACoS Moved

The market comparison tells you whether you drifted more than everyone else. It cannot tell you where. If your increase is bigger than the market's, the cause usually sits in a few SKUs losing the Buy Box, a competitor bidding up your branded terms, or budget caps cutting off your best campaigns in the afternoon. Finding it means joining spend to sales at the SKU and search-term level every week.

That is where Qore, the workflow layer from Trellis, fits. Describe the review you run (which SKUs, which thresholds, what counts as drift), and Qore turns it into a scheduled workflow that checks your account data and returns the SKUs and campaigns where TACoS moved more than the market did. Proposed changes wait for your approval by default, and you decide which workflows can act on their own. Qinetix, our ads automation platform, then manages bids and budgets against your TACoS target within the guardrails you set. Qore is in open beta, and you can sign up here.

Conclusion: Plan for a More Expensive Auction, and Measure Against It

Amazon's own numbers show ad spend taking a growing share of marketplace sales, by roughly half a point to a point a year. Part of that is new ad inventory rather than higher prices for sellers, and the exact level is an estimate. The direction is not in doubt.

Three things follow. Compare your TACoS change to the market's before deciding something broke. Build some drift into next year's plan. And put your spend behind the products whose margins can carry a more expensive auction. If your drift turns out to be bigger than the market's, the next question is which SKUs.