Why your Amazon ad data does not match your sales

During an account audit, Alfredo Roselli found a discrepancy that looked impossible: the advertising console credited more ad-attributed units than the business had sold in total for the same period. The answer was not a magic reporting fix. The team had to separate attributed conversions from actual orders and then inspect how the campaigns were eligible for credit.

If your ad dashboard looks better than your P&L, start with the same question: what exactly is this report counting, and which interaction made the conversion eligible?

Attribution is a measurement rule

Amazon defines attribution as a conversion that occurs after an eligible ad interaction. The interaction, product relevance, campaign type, and lookback window determine whether a campaign can receive credit. A report date can remain incomplete until its lookback window closes, and conversions can appear later while being reflected on the date of the ad interaction.

That means an attributed purchase is not the same thing as an incremental purchase. Attribution tells you which campaign received platform credit. Incrementality asks whether the purchase would have happened without that ad. Use both ideas, but do not treat one as the other.

Lookback windows vary by campaign and advertiser type

There is no single Amazon-wide attribution window. Amazon's current campaign-attribution documentation lists Sponsored Products seller attribution as a seven-day click-only window. For vendors and authors, the same table lists a 14-day click-only window. View-enabled campaign types can use view attribution, and their eligibility depends on the campaign and cost type.

Check the definition in the report you are using. Amazon's reporting documentation also distinguishes click-attributed metrics from all-view metrics. A dashboard that combines them can look very different from a business report for the same dates.

Two account-level checks Enflet uses

1. Confirm the seller who captured the order

Sponsored Products generally depend on a product offer that can receive the order. Sponsored Display can create a different diagnostic when the advertised listing is not controlled by your offer. If your ad sends a shopper to a detail page while another seller owns the Featured Offer, check the offer and order records before treating the attributed sale as your revenue.

This is an account-level hypothesis to validate, not a universal behavior claim. Compare the ASIN, seller, offer status, and order data for the exact dates and campaigns. If you paid for the traffic but another seller captured the order, the advertising report needs a more careful interpretation.

2. Separate click and view credit

Amazon says some campaign types can attribute conversions to clicks and views, depending on the campaign and cost type. Sponsored Display campaigns billed on a viewable-impression basis are among the campaign types affected by view-attribution reporting updates. A shopper may see one ad, later click another, and then buy. The reports can expose both interaction types when the campaign rules allow it.

Export click-only and view-inclusive metrics where available. Do not add every campaign's attributed sales together and call the result incremental revenue. Use the report dimensions and campaign definitions to identify overlap.

How to reconcile the dashboard with the business

• Choose one period, marketplace, and time zone. Keep the dates identical in every report.
• Start with Amazon business reports or your order and P&L system for units and revenue.
• Pull advertising data by campaign, advertised ASIN, targeting, interaction type, and report date.
• Label click-attributed, view-attributed, promoted-product, and brand-halo metrics separately.
• Check Featured Offer status, seller identity, cancellations, returns, and refunds for ASINs with large gaps.
• Wait for the relevant lookback window to close before judging the final total.

Then ask the business question: did the ads create profitable demand, or did the platform assign credit to a purchase that another campaign, another seller, or existing brand demand already influenced?

What the discrepancy does not prove

A mismatch does not prove that Amazon Ads is broken or that Sponsored Display is useless. Sponsored Display and other view-enabled formats can support discovery and consideration. It does prove that one blended ACOS or ROAS number is too small a lens for an account-level decision.

Use attributed metrics for campaign management. For a companion on the ratio itself, see what a good ACOS means. Use total orders, contribution margin, organic trends, and controlled tests to evaluate business impact. If the attribution rules changed or a report is still inside its lookback period, mark the result as provisional.

Common questions

Why does Amazon ad data not match my sales?

Common causes include different date or marketplace filters, delayed conversions, campaign-specific lookback windows, product or brand-halo eligibility, view attribution, and a seller other than you capturing the order. Reconcile the same ASINs and dates across business and advertising reports.

What is view-through attribution?

It is a conversion credited after a shopper sees an eligible ad without clicking. Whether it applies depends on the campaign type and cost model. View-attributed metrics should be read separately from click-only metrics.

Is Amazon attribution the same as incremental sales?

No. Attribution is platform credit under defined rules. Incremental sales are purchases that would not have happened without the advertising. Attribution can inform a test, but it cannot prove incrementality by itself.

Which attribution window should I use?

Use the window Amazon defines for the campaign and advertiser type in the report. The current official table lists seven days for seller Sponsored Products clicks and 14 days for vendor and author Sponsored Products clicks. Other campaign types and view reporting can differ.

If your dashboard and business numbers disagree, Enflet's Amazon PPC management team can help trace the gap from campaign interaction to order and margin.