Fabric and Azure
Tableau to Power BI Migration: What Transfers
A Tableau to Power BI migration means rebuilding each view by hand. Here is what transfers, why totals change, and how to check numbers before cut-over.
Austin Levine
·
Aktualisiert
A Tableau to Power BI migration moves your dashboards, data connections, and calculations from Tableau into Power BI. Workbooks don't convert on their own. You rebuild each view, translate calculated fields into DAX measures, and repoint data sources. Plan four to twelve weeks for a mid-size estate, and validate every number against the old reports.
Most teams treat a Tableau to Power BI migration as an export job, and it shows up at the first month-end when a revenue figure on the new report sits three percent off the Tableau number leadership trusted for years.
That gap is the part that matters. A report that looks identical to the old one but returns a different total will cost you the trust of the people reading it. So before you move a single workbook, it pays to know what carries over on its own, what gets rebuilt by hand, and where the totals drift without anyone noticing.
Here's the honest version. The two platforms store data, build calculations, and draw charts in different ways, so a migration is closer to a controlled rebuild than a file transfer. Power BI now runs in more than 375,000 organizations, and Microsoft has ranked as a Leader in the Gartner analytics Magic Quadrant for 18 years running, so the destination is proven. Getting there cleanly is the work.
Key takeaways
A Tableau to Power BI migration moves dashboards and connections, but workbooks do not convert on their own.
You rebuild each view, translate calculated fields into DAX measures, and repoint data sources.
Plan four to twelve weeks for a mid-size estate and validate every number against the old reports.
Four settings cause almost every number mismatch, and each one is fixable.
What actually transfers in a Tableau to Power BI migration?
Short answer: less than you'd hope, and that's fine once you plan for it. Your data connections point at the same databases, so those move with light editing. The visuals, the calculations, and the security all get rebuilt inside Power BI.
Here's how the main pieces map across.
Tableau piece | Power BI equivalent | Moves on its own? |
|---|---|---|
Workbook / dashboard | Report (.pbix) | No, rebuilt by hand |
Data source connection | Semantic model + Power Query | Partly, repointed |
Calculated field (row) | Calculated column | No, rewritten in DAX |
Level-of-detail / aggregate field | Measure | No, rewritten in DAX |
Data blend | Model relationship (star schema) | No, remodeled |
Published access rights | Workspace roles + row-level security | No, reset |
Extract refresh schedule | Scheduled or incremental refresh | No, reconfigured |
The pattern is consistent. Anything that's a pointer to data moves with small changes. Anything that carries logic, a calculated field, a blend, an access rule, gets recreated in Power BI terms. Tableau workbooks have no import path into a .pbix file, so every dashboard is rebuilt view by view.
Why do the numbers change after the move?
Four settings cause almost every mismatch, and each one is fixable once you know which one you're looking at.
First, calculation context. Tableau uses level-of-detail expressions and table calculations that recompute against whatever is on the sheet. Power BI uses DAX, where a measure recalculates against filter context. A field that summed correctly in Tableau can double count in Power BI when the model has a many-to-many relationship you didn't expect.
Second, default aggregation. Tableau treats most numeric fields as SUM by default. In Power BI you write explicit measures, so a field left as an implicit aggregation can total the wrong way without warning.
Third, data blending. Tableau blends two sources on a shared field at the sheet level. Power BI wants a proper star schema with defined relationships, so a blend that averaged fine can shift once the same data sits in modeled tables.
Fourth, date handling. Tableau builds date hierarchies for you. In Power BI you need a dedicated date table marked as a date table, or time intelligence like SAMEPERIODLASTYEAR returns blanks.
Given that 67% of organizations already say they don't fully trust the data they use to make decisions (Precisely and Drexel LeBow, 2025), a new report that disagrees with the old one by even a small amount undoes the reason you moved. Validation is the step you can't skip.
How do you rebuild Tableau calculated fields in Power BI?
Do it in three passes rather than one, so you catch the logic before you polish the layout.
Inventory every calculated field. Export the Tableau workbook to a .twb file, which is XML, and pull the calculation list, or read them from the data source pane. Sort them by how often the dependent sheets are viewed, using the usage stats in Tableau Server or Tableau Cloud, so the busiest calculations get attention first.
Sort each one into row level or aggregate. A row-level calculation that runs per record becomes a calculated column. An aggregate calculation, or a level-of-detail expression, becomes a DAX measure. This one split prevents most double-counting later.
Rewrite and test in pairs. Put the Tableau result and the Power BI measure side by side for the same filters, and match them to the cent before you move on. A measure like `Sales LY = CALCULATE([Total Sales], SAMEPERIODLASTYEAR('Date'[Date]))` should return the identical figure your Tableau expression returned for last year.
What about the data source, and do you need to migrate to Azure SQL?
Your connections are the easiest part to move and the easiest to over-scope. When Tableau reads from SQL Server, Power BI reads from the same SQL Server with a new connector, and you decide between Import mode and DirectQuery per report.
A migration is also a natural moment to review where the data lives. Many teams choose to migrate to Azure SQL at the same time, so the reports and the database move to Azure together and refresh stays fast. When you run that step, treat it as its own project: you migrate SQL to Azure SQL first, confirm every table and row count matches, and only then repoint Power BI at the new endpoint. Doing both moves in one weekend is how a query that ran in two seconds starts timing out in front of the finance team on Monday.
For large tables, set up incremental refresh with RangeStart and RangeEnd parameters, so Power BI only pulls the newest rows on each refresh instead of reloading years of history. That one setting is the difference between a five-minute refresh and a five-hour one.
How long does a Tableau to Power BI migration take?
For a mid-size estate of 30 to 60 dashboards, plan four to twelve weeks with one or two developers. The spread depends less on the visuals and more on how much hidden logic sits inside the Tableau calculations. A workbook with 200 calculated fields takes longer than ten simple dashboards.
Run it in phases so value lands early:
Week 1: inventory and rank every workbook by views, and pick the top ten that carry the real decisions.
Weeks 2 to 4: rebuild the data model as a star schema and translate the calculations.
Weeks 4 to 8: rebuild visuals, apply row-level security, and validate every number against the retiring Tableau report.
Weeks 8 to 12: user testing, then cut over.
Use Power BI deployment pipelines so the same report promotes through development, test, and production without manual re-uploads, and set row-level security roles early so you test them with View As before anyone sees live data.
Frequently asked questions
Does Power BI have a tool to convert Tableau workbooks automatically? No supported converter turns a .twb or .twbx into a working .pbix. Third-party tools can read the Tableau XML and list your calculations, which speeds up the inventory, and then a person rebuilds each report in Power BI.
Will my Tableau level-of-detail expressions work in Power BI? Level-of-detail expressions become DAX measures that use CALCULATE and the right filter functions. You rewrite the logic and validate each measure against the Tableau output before you trust it.
Can I move Tableau and my database to Azure at the same time? You can, and many teams do, though it pays to sequence it. Migrate the database first, confirm the row counts, then repoint Power BI. Changing the reports and the data source in one step makes any mismatch hard to trace.
How do I keep the old Tableau reports running during the migration? Run both in parallel until every number matches. Keep Tableau live, publish the Power BI version to a private workspace, and let a few analysts compare them for a full reporting cycle before you switch the audience over.
What's the single biggest mistake in a Tableau to Power BI migration? Rebuilding the visuals before the data model. Teams recreate the charts first, then discover the totals are off, and end up redoing the work. Model the star schema and the measures first, prove the numbers, then make it look good.
Read next: Power BI migration: the full playbook and Migrate SQL to Azure SQL: the real checklist.
