Automate Amazon Ads reports before the old report pages retire on 31 December 2026

If your team starts the week with Amazon Ads reports emailed from the advertising console, those reports have a deadline. In June 2026 Amazon Ads announced that the Sponsored Ads reports and Amazon DSP reports pages will be retired by 31 December 2026, in favour of a new unified reporting page. Scheduled reports set up on the old pages need to be moved or rebuilt before then.

That makes this a good moment to automate Amazon Ads reports properly, so the numbers land in a sheet, a dashboard or your own database every morning without anyone downloading a file. This post is for sellers and brands running Sponsored Products who either pull reports by hand or depend on scheduled reports they have not looked at in a year.

I build Amazon PPC automation, including rules for bids, negative keywords and budgets, and the reports behind them. Below is what is changing, the options for replacing old reports, and how an automated pipeline works.

What retires on 31 December 2026

Amazon's June 2026 announcement names two pages in the Ads Console: Measurement and reporting > Sponsored Ads reports, and Measurement and reporting > Amazon DSP reports. Both are being sunset by 31 December 2026. Their replacement, unified reporting, sits under Measurement and reporting > Reporting and became generally available in the same announcement, after an open beta that began at unBoxed in November 2025.

What unified reporting offers, according to Amazon:

• custom reports or pre-built templates, which can be scheduled;
• one report spanning several accounts, ad products, countries, metrics and dimensions;
• standardised metrics and dimensions, with traffic-based attribution across ad products;
• up to 15 months of daily or weekly data, and up to six years of monthly, yearly or summary data.

For scheduled reports, Amazon said an automated migration tool would be "available soon". Check whether it has reached your console. Either way, compare the migrated reports with the originals, because standardised metrics may not match old column definitions exactly.

Three ways to replace your old reports

There are three sensible paths, and many sellers mix them.

Option	Good for	Watch out for
Recreate schedules in unified reporting	A person reads a weekly file	Data stays in files and inboxes
A third-party connector to Google Sheets or Looker Studio	Quick charts without code	A monthly fee and the vendor's schema
Your own pipeline on the Amazon Ads API	Joining ads data with sales, stock and costs	Someone has to build and maintain it

If the report is read once a week by one person, recreating it in the console is enough. If the numbers feed decisions every day, or need to sit next to sales and stock, automation pays for itself in hours saved and mistakes avoided.

How to automate Amazon Ads reports with the API

Custom pipelines run on the Amazon Ads API. Amazon describes it as a way to report on advertising activity "in a programmatic and scalable way", and its API page says direct advertisers can apply for access and that there are "no additional fees from Amazon Ads to use the API".

Reporting through the API is asynchronous. A nightly job works like this:

1. Request. Ask for a report: the report type (campaigns, targeting, search terms, advertised products), the columns, the date range and whether you want daily rows or a summary.
2. Wait. Amazon builds the report in the background. The job checks the status until it is complete.
3. Download. Fetch the file, parse it and write the rows to your database or sheet.
4. Repeat a window. Re-pull a trailing window of recent days on every run, not only yesterday, so late corrections are picked up, and overwrite rather than append.
5. Alert on failure. If a report fails or comes back empty, someone should hear about it that morning, not at month-end.

Two practical points. First, keep the raw files: when you change a calculation, you can rebuild history instead of re-requesting it. Second, keep the reporting code behind one small module. Amazon's June notice lists the Reporting API as one of the ways to reach unified reporting, still in beta, so the request format may change. A pipeline built that way swaps one module, not everything.

Hourly data with Amazon Marketing Stream

Daily reports suit most decisions. When you need to act within the day, Amazon Marketing Stream is the other route. It is push-based: Amazon sends hourly performance metrics and campaign changes for sponsored ads to your own AWS account, through Amazon SQS or Amazon Data Firehose, so you do not poll for them. It needs an Amazon Ads API integration and an AWS account, and Amazon's June notice lists unified reporting through Marketing Stream as beta too.

Typical uses are an alert when a profitable campaign runs out of budget by mid-afternoon, or a few weeks of hourly data to see whether dayparting is worth it. If nobody will act on an hourly signal, skip it.

Where the data should land

The same pipeline can feed several places:

• A Google Sheet for a small account: one tab of daily rows per report, with summary tabs on top. Simple, familiar and easy to share.
• A database behind a dashboard once you want ads next to everything else. Spend and ad sales come from the Ads API; total sales, stock and fees come from Seller Central through the Selling Partner API, covered in automating Seller Central. Together they give total ACOS, often called TACOS (ad spend as a share of all sales, not only ad-attributed sales), and profit after ads per product.
• A morning message by email or Slack: yesterday's spend, sales and ACOS by product, plus anything that crossed a threshold.

If you sell on more than one channel, the ads feed is one input into a wider view. I cover the data model and which numbers to show in one sales dashboard for every channel.

From the search term report to negative keywords

The search term report is the most useful report to automate, because it drives action. It shows the actual searches that led to clicks on your ads. A nightly job can read it, find terms that have spent past your limit without an order, and propose them as negative keywords. The same pass can find converting search terms in automatic campaigns and propose them as exact-match keywords in a manual campaign.

Keep a person in the loop. A short approval queue each morning is safer than a rule that silently blocks a search you later want back. The thresholds, guardrails and rule design are covered in my guide to Amazon PPC automation, so I will not repeat them here.

A checklist before the old pages close

1. List every scheduled report on the old pages: who receives it and which decision it supports.
2. Drop the ones nobody reads. Retirement is a free clean-up.
3. Recreate or replace the rest, in unified reporting or as an API feed.
4. Run old and new side by side for the same date range and compare spend, sales and clicks. Investigate any gap before you switch.
5. Download history you want to keep in its current form before the pages go.
6. Set alerts for failed or empty runs.

Do this in November, not in the last week of December, when peak-season campaigns leave no time for surprises.

Where AI fits

Since 2 February 2026 the Amazon Ads MCP server has been in open beta for partners with Ads API credentials. It lets AI assistants such as Claude, ChatGPT or Gemini run Ads API functions, including reports, from a plain-language request. That is handy for one-off questions ("which campaigns lost money last week?"). For reports that must arrive every morning in the same shape, plain scheduled code is cheaper and more predictable.

What automated ads reporting costs

Unified reporting is free, and connectors cost a monthly subscription. A custom pipeline to automate Amazon Ads reports costs more up front, and the data and logic are yours.

A daily ads report into a sheet or a simple dashboard is the kind of single automation my Marketplace & e-commerce apps MVP plan covers, from $2,500: one marketplace through its official API, one automation such as reports. Combining ads data with sales and stock from two or more channels, with a dashboard, starts from $5,500 on the Flexible plan. These are starting prices; every project gets a fixed quote after a short brief, and you own the code. More on how ads fit with the rest of a seller's setup on my Amazon PPC automation page.

FAQ

Will my scheduled Amazon Ads reports stop after 31 December 2026?

Reports scheduled on the Sponsored Ads reports and Amazon DSP reports pages depend on those pages, which Amazon said in June 2026 it would retire by 31 December 2026. Recreate them in unified reporting under Measurement and reporting > Reporting, or use the migration tool Amazon said would be available soon. Then compare the new versions with the old ones for the same dates before relying on them.

Do I need a developer to get Amazon Ads data into Google Sheets?

Not always. Unified reporting can schedule reports, and several third-party connectors pull Amazon Ads data into Google Sheets or Looker Studio for a monthly fee. A developer makes sense when you want ads data joined with sales, stock and costs from Seller Central, kept in your own database, or used to drive rules such as negative-keyword proposals rather than only displayed.

What is the difference between the Reporting API and Amazon Marketing Stream?

The Reporting API works on request: your tool asks for a report covering a date range, waits while Amazon builds it, then downloads the file. Marketing Stream pushes hourly performance metrics and campaign changes to your own AWS account through Amazon SQS or Amazon Data Firehose. Most reporting needs only the Reporting API. Stream earns its place when you act within the day, for example with hourly budget alerts.

If you want your ads reports rebuilt before the deadline, send me a short brief: which reports you schedule today, who reads them, where the numbers should land and which marketplaces you advertise in. For the wider picture, start with e-commerce automation: what to automate first.