What Power BI on Business Central actually looks like: the extraction patterns, the phase-by-phase timeline, and how Hopton's approach compares to Jet and Cosmos.
If you run Business Central, at some point someone in Finance or Ops asks for a dashboard, and the native reporting doesn’t quite get there. This is what actually happens once you decide to fix that properly, not the marketing version.
The simplest route is Power BI connecting directly to BC through the standard connector, with no separate warehouse or lakehouse in between. Power BI pulls data via the BC APIs, models it, and serves reports. For straightforward reporting needs, that simplicity is the right answer and there is no reason to overbuild.
The harder cases are the ones with real history behind them: multi-company consolidations, currency translation, year-end balances, dimension breakdowns. BC extraction has a well-documented set of gotchas that catch people who haven’t hit them before. Posting groups affect how transactions roll up in ways that are easy to miss. BC’s analytical dimensions need careful translation into report-friendly structures. Closing entries at year-end can look like duplicates if they aren’t handled properly, and currency revaluations distort historical comparability if nobody accounts for them. None of this is exotic, but it is invisible until you’ve been burned by it once.
This is also why schema stability matters more than people expect. BC’s data structure changes meaningfully between versions, especially around major Wave releases, and extraction built against one version can need adjustment when the source upgrades. A layered Bronze, Silver, Gold architecture absorbs those changes in the data layer, so reports on top stay stable and the maintenance work never touches what users see. That protection is one of the strongest arguments for building it this way rather than bolting reports directly onto raw BC tables.
Two named products come up constantly when BC customers look beyond Microsoft’s own tools: Jet Reports, now under insightsoftware, mature and Excel-based with deep roots in the BC partner channel; and Cosmos, a newer cloud-only, BC-only entrant with a notably fast install. Both are credible, and both have real limits. Cosmos suits organisations that want operational BC reporting quickly with minimal setup, but it locks you into a single vendor’s analytical model and doesn’t extend to multi-source analytics or on-premises BC environments. Jet still makes sense for BC-only, Excel-first finance teams who value that heritage. Hopton’s approach instead builds on the Microsoft stack directly, Fabric, Power BI, the Lakehouse Connector and Copilot, which means no third-party dependency and an architecture that sits on the same platform as the rest of your Microsoft 365 investment. It’s more upfront work than a packaged tool, and it earns that cost back when you have multi-source needs or AI ambitions that a BC-only product can’t reach.
A typical Business Central on Power BI implementation runs about twelve weeks. Weeks one to three cover discovery and architecture: auditing BC, agreeing priority reports, and designing the Bronze, Silver, Gold structure. Weeks four to eight are the build itself, extraction, the Silver layer, the Gold layer, semantic models and the priority reports. Weeks nine to eleven are testing and refinement, reconciling against existing BC reports and tuning performance, and week twelve is go-live and handover. Throughout that period, BC operations and any existing reporting tools keep running unchanged. The new analytics layer goes live alongside what you already have, and old tools get retired deliberately once the new one has proven itself, not by accident on day one.
A question we hear early in almost every conversation is whether we can do this without the client’s existing BC partner being involved. We can. The BC partner relationship covers the ERP itself; the analytics layer sits alongside it and doesn’t require renegotiating that relationship or routing decisions through a third party who may have no particular expertise in Power BI or Fabric.
Craig Daniels
Senior Analytics Solutions Consultant
Part of the Hopton Analytics team, delivering governed analytics programmes for UK mid-market organisations.