Amazon Seller & Vendor Central to Power BI

Connect Amazon Seller Central and Vendor Central to Power BI to compare sales, stock and margin across your marketplaces. Start with the reports your account can access, then use a Selling Partner API (SP-API) pipeline or a third-party connector to stage them for Power BI. Microsoft lists connectors for other Amazon services, such as Athena, but these are different from a Seller Central or Vendor Central connection. A warehouse is useful when you need to preserve report history or combine several accounts. Microsoft's connector catalogue identifies the available Power BI sources.

Seller Central, Vendor Central and DSP: name your programme first

The method and the metrics change depending on which programme you run, so name yours before you pull a single report.

• Seller Central is for third-party sellers who sell directly to customers. Select the orders, finance, inventory and report data your seller account is authorised to access through SP-API.
• Vendor Central is for suppliers who sell stock to Amazon. Its retail analytics reports cover sales, traffic and inventory; these describe a different business relationship from seller orders. Amazon's analytics report definitions specify availability, roles and report options.
• Amazon DSP is a separate advertising programme. Its campaign reporting uses the Amazon Ads API. Confirm that a connector supports the DSP reports you need as well as any Sponsored Ads reports.

Many businesses run more than one programme. Agree the reconciliation between seller, vendor and advertising data so that revenue and ad costs are each counted once.

Knowing your programme tells you which reports to pull and which KPIs make sense.

Advertising is a separate source. Use the Amazon Ads API for campaign reporting. Confirm access to both services if you want ad spend beside marketplace sales. An advertising attribution total represents credited sales; adding it to retail sales would count some purchases twice.

Three ways to connect Amazon to Power BI

The Selling Partner API gives you control over orders, finances, inventory and reports available to your account. A cloud pipeline or script can stage the results in a database. Account authorisation, report generation and throttling are part of the extraction work. Keep the Amazon Ads connection separate.

A third-party connector removes most of that plumbing. Several vendors publish Amazon connectors that handle authentication and incremental loads, and you point Power BI at their service or staged tables. This is the quickest route to a working dashboard, but you take on a subscription and a dependency on someone else's roadmap.

A done-for-you build lands Amazon data in a warehouse you own, then models it in Power BI. This is the right pattern when you sell across several marketplaces, need clean history, or want sales joined to costs and ad spend. See our data integration work for this route.

A report-to-dashboard acceptance checklist

Use this reference checklist to agree the build before connecting every marketplace:

1. Record the source contract. Name the programme, account, marketplace, report type, reporting period and currency. For vendor sales, also record the chosen distributor view and selling programme. Keep each extract with its request parameters.
2. Wait for a completed report. A report request starts a job. Retrieve the document after completion and retain its identifier with the loaded rows. Amazon's report tutorial describes the request, completion notification and document download sequence.
3. Keep each table's grain explicit. A product-by-day retail total and an order-line transaction need different tables. Join shared product and date dimensions to those tables rather than multiplying totals through a direct many-to-many join.
4. Reconcile like with like. Compare a vendor sales extract with the same vendor report and options. Reconcile seller cash movements separately to the relevant finance or settlement report. Compare advertising with its own reporting window.
5. Test a repeated load. Reload the same completed report. The accepted result is unchanged row counts and totals, with the second run recorded in the load log.

Illustrative reconciliation example: a vendor report contains two product rows with 12 and 8 shipped units for one period and marketplace. The matching Power BI view should show 20 shipped units with those same filters. Changing the view to ordered units creates a different check. This is a model test, not a client result or a revenue forecast.

Marketplace KPIs worth modelling

Once the data flows, the value is in the measures. Useful Amazon KPIs include net revenue after fees, units sold, average selling price, Amazon fees as a share of revenue, advertising cost of sales, return rate, buy box share, inventory cover and contribution margin per product. Model these by marketplace and product so you can see which lines actually make money after fees. A clean date dimension and an agreed settlement-versus-order date rule matter more than any clever calculation, because Amazon's own reports shift as data settles.

Common pitfalls

• Eventual consistency. Amazon reports update as data settles. A figure pulled today can change tomorrow. Reconcile to settlements and refresh on a sensible schedule.
• Date logic. Order date, ship date and settlement date all differ. Decide which drives revenue and document it.
• Fees and returns. Net profit needs fees, refunds and reimbursements modelled correctly, not just gross sales.
• Throttling. The SP-API rate-limits requests. Use incremental, report-based loads rather than pulling everything each time.
• Multi-marketplace currency. Standardise on one reporting currency and record where conversion happens.

Choosing the right route

For one marketplace and standard reports, compare a connector's coverage with the tables you need. For several marketplaces, a hybrid Seller/Vendor business or retained report history, assess a warehouse-backed model. Include extraction, monitoring and source maintenance in the scope. Our published BI pricing explains the review and build stages; the US and Australia pages cover delivery in those markets.

If you need seller, vendor and advertising figures in one reporting model, a Power BI consultant can define the separate tables, shared dimensions and reconciliation checks before the dashboard build.

Start with a Trusted Numbers Review to examine the reporting you have and get written findings and a prioritised plan. Bring a sample report, its filters and the decision you want the dashboard to support.

For a first pass at your reporting process, use the free two-minute reporting checklist. Its result reflects your answers; a review tests the reports and data.

Technical sources checked on 6 September 2026. Confirm report access and options for your account when scoping the connection.