Amazon Attribution Windows: What They Measure, What They Miss, and How to Adjust

Every Amazon ad report you pull is shaped by one setting most sellers never think about: the attribution window. It determines how long after a click Amazon gives credit to your ad for a sale. Change the window, and the same campaign can look profitable or unprofitable depending on which number you are reading.

Most sellers accept the default settings and never question them. That is a mistake, because the defaults are not neutral. They shape how you evaluate performance, where you allocate budget, and which campaigns you scale or cut. Understanding the mechanics is not optional if you are spending real money on Amazon ads.

How Amazon Attribution Windows Work

When a shopper clicks your Sponsored Products ad and buys your product three days later, Amazon attributes that sale to the ad. The "window" is how many days after the click Amazon keeps looking for a purchase to credit.

The defaults vary by ad type:

• Sponsored Products: 7-day click attribution. A sale must happen within 7 days of the click to count.
• Sponsored Brands: 14-day click attribution. Twice the lookback window of Sponsored Products.
• Sponsored Display: 14-day click attribution, plus 14-day view attribution (someone saw your ad but did not click, then bought within 14 days).
• Amazon DSP: 14-day click attribution and 14-day view attribution by default, configurable up to 28 days for both.

These are not arbitrary numbers. Amazon chose them because they roughly match the purchase decision timeline for most consumer products. But "most" is doing a lot of work in that sentence. If you sell a $12 phone case, the purchase decision takes minutes. If you sell a $400 baby stroller, shoppers research for weeks. The same attribution window cannot accurately represent both.

Why the Default Windows Distort Your Data

Sponsored Brands Gets Inflated Credit

Sponsored Brands has a 14-day window while Sponsored Products has 7 days. That means Sponsored Brands captures sales that happened up to two weeks after the click, while Sponsored Products stops counting after one week. Run both campaign types on the same product and Sponsored Brands will almost always show a better ROAS, not necessarily because it performs better, but because it has a longer window to collect conversions.

We see this constantly when auditing accounts. Brands pour budget into Sponsored Brands because the ROAS looks 30% to 50% higher than Sponsored Products. Some of that is real (Sponsored Brands does drive brand awareness that converts later), but some of it is just the math of a longer window.

DSP View-Through Attribution Inflates Everything

DSP counts a sale as attributed if someone merely saw your ad and then purchased within 14 days, even if they never clicked. On a product with strong organic sales, a significant chunk of those "attributed" DSP sales would have happened anyway. The shopper was already going to buy your product. They happened to see a display ad during their normal browsing, and DSP claimed credit.

This does not mean DSP is worthless. It means you need to look at incrementality, not just attributed ROAS. The question is not "did DSP touch these sales?" but "would these sales have happened without DSP?" Those are very different questions, and the attribution report only answers the first one.

Short Windows Miss Long Consideration Cycles

For products with purchase cycles longer than 14 days (furniture, electronics, specialty equipment), even the longest attribution window misses real conversions that your ads influenced. A shopper clicks your Sponsored Products ad, researches for three weeks, and comes back to buy. That sale shows up as organic in your reports, and you might cut the campaign that drove it because the attributed ROAS looks weak.

How to Adjust Your Approach

Compare Apples to Apples

When evaluating campaign types against each other, normalize the attribution window. Most Amazon ad consoles let you toggle between 7-day and 14-day views. Compare Sponsored Products and Sponsored Brands both on a 7-day window to get a fairer comparison. The Sponsored Brands numbers will drop, but you will get a more accurate picture of relative performance.

Track Total Sales Alongside Attributed Sales

Pull your total revenue for each ASIN alongside the attributed revenue from ads. Calculate your TACoS (Total Advertising Cost of Sales) as a percentage of total revenue, not just ad-attributed revenue. If you increase ad spend and total revenue grows proportionally, the ads are working even if the per-campaign ROAS looks flat. Attribution windows create noise at the campaign level, but TACoS smooths it out at the product level.

Use Amazon Marketing Cloud for Multi-Touch Analysis

If you are running DSP, you have access to Amazon Marketing Cloud (AMC). AMC lets you see the full path to purchase: which ads a shopper saw, in what order, and how long the journey took. This is the closest Amazon gets to true multi-touch attribution instead of last-touch. It is not simple to set up or query, but for brands spending $20,000+ per month on ads, the insights justify the effort.

Run Incrementality Tests

The most reliable way to measure real ad impact is controlled testing. Turn off DSP for a specific product or audience segment for two weeks and measure the change in total sales. If sales drop by the amount DSP was claiming as attributed, it was genuinely incremental. If sales barely move, DSP was taking credit for organic demand.

We run these tests quarterly for our clients. The results vary widely by product and category. Some brands see 60% to 70% true incrementality from DSP. Others see 20%. The only way to know your number is to test.

Build Your Budget Around TACoS, Not Campaign ROAS

Attribution window quirks make campaign-level ROAS an unreliable signal for budget decisions. A Sponsored Products campaign showing 3x ROAS on a 7-day window might actually be driving 5x total return when you account for the organic sales it influences beyond the attribution cutoff. A DSP campaign showing 8x ROAS on a 14-day view-through window might only be driving 2x incremental return.

TACoS captures the full picture. If your total ad spend is $30,000 and total revenue is $300,000, your TACoS is 10% regardless of how individual campaigns attribute. Month over month, if TACoS is stable or declining while revenue grows, your advertising is working. That is a more reliable signal than any single campaign report.

The Bottom Line

Attribution windows are a measurement tool, not a performance tool. They tell you where Amazon decided to assign credit, not where the actual influence happened. Smart budget decisions come from understanding what the numbers include, what they miss, and building a reporting framework that accounts for both.

If your Amazon advertising decisions are based solely on campaign-level ROAS from the default reports, you are almost certainly misallocating budget. Not dramatically, but enough to leave meaningful growth on the table. Understanding attribution is one of the fastest ways to improve ad efficiency without changing a single bid.