Illustration of a laptop showing bar charts and line graphs on a dashboard

Moving insurance reporting from Excel to Power BI

  • Darren Vinales
  • October 19, 2026

Excel is still where most insurance reporting starts. Claims listings, premium summaries, reserve movements and board packs are often built in spreadsheets that have grown over several years. They work, but they tend to depend on a few people, on manual copying of data and on careful checking every period. Power BI offers a way to keep the logic you already have while making the reports easier to refresh, share and control.

This article sets out when a move makes sense and how to approach it.

Signs that a spreadsheet report has outgrown Excel

  • The same steps are repeated every period. Data is exported, pasted into a workbook, reformatted and checked by hand each month or quarter.
  • Several versions of the truth. Different teams keep their own copies, and figures for the same measure do not always agree.
  • Workbooks are slow or fragile. Large claims or policy files make the workbook hard to open, and a change in one tab can break another.
  • Only one person understands it. The report relies on knowledge that is not written down.
  • Users want to explore the figures. Management asks follow-up questions that need a new version of the spreadsheet each time.

What Power BI changes

Power BI separates three things that are often mixed together in a spreadsheet: getting the data, modelling it and presenting it.

  • Getting the data. Power Query connects to your systems and files and records each cleaning and reshaping step, so the same steps run again on refresh.
  • Modelling the data. Tables such as policies, claims and payments are linked by relationships rather than lookups, and measures are defined once and reused across reports.
  • Presenting the data. Reports are published to a shared service, so users see the same version and can filter and drill into the figures themselves.

How to plan the move

  1. Start with one report. Pick a report that is produced regularly and causes the most manual effort, such as a monthly claims summary.
  2. Write down the logic. Before building anything, document where each figure comes from and how it is calculated. This is often where differences between versions come to light.
  3. Connect to the source, not to the spreadsheet. Where you can, connect Power BI to the underlying system or extract rather than to the old workbook, so the new report does not inherit its manual steps.
  4. Run old and new side by side. Produce both for at least one reporting cycle and explain any differences before retiring the spreadsheet.
  5. Agree who owns it. Decide who maintains the data connections, the model and the report layout, and make sure they are trained.

Common pitfalls

  • Rebuilding the spreadsheet as it was. A direct copy of a crowded workbook rarely makes a good dashboard. Ask users which questions the report needs to answer.
  • Ignoring data quality. Power BI will show data issues more clearly, not fix them. Build checks into the data preparation steps.
  • Leaving out access control. Claims and policyholder data can be sensitive, so decide early who should see what.

Excel still has a place

Moving to Power BI does not mean giving up Excel. Power Query and Power Pivot work inside Excel too, and are a good way to start automating data preparation before a full move. Our article on Power Query and Power Pivot for actuarial and finance teams covers this in more detail.

How we can help

We build Power BI reporting for insurers as part of our business intelligence reporting and analytics service, and the same data processes can support reserve reviews and regulatory returns . To discuss your reporting, please contact us .

Related Posts

Illustration of a desk with a calculator, a laptop and data charts on transparent screens

Power Query and Power Pivot for actuarial and finance teams

  • Darren Vinales
  • October 19, 2026

How insurance actuarial and finance teams can use Power Query and Power Pivot in Excel to automate data preparation, build models and cut manual work.

Read More