The Rows Multiplied After the Join. Did the Business Really Grow?
After a Power Query join, check the intended relationship, row counts and recognisable records before interpreting a larger total as business growth.

You combine the sales file with a lookup table. The report now contains more rows, and the total is larger. Before celebrating growth, find out what one row in the combined result actually represents.
A join connects matching records. If the relationship is different from the one you expected, expanding those matches can change the number of rows in a way that needs explanation.
Define the expected relationship
Consider a fictional small wholesaler. Its transaction file contains order lines, while a product lookup should describe each product once. The coordinator expects each line to find one product description.
If the lookup contains several records for a product code, a line can match more than one record. That may be a data issue, or the lookup may legitimately describe versions or locations. The reporting decision depends on the intended relationship.
Microsoft’s Power Query guidance explains that a merge joins tables using selected matching columns. Choose those columns and the join type deliberately. Do not treat a familiar-looking dialog as evidence that the business relationship is correct.
Count at the useful checkpoints
Record the starting row count, the expected matches and the count after expansion. Also compare a relevant amount where appropriate. Explain any intended increase, such as an order being represented by its separate lines.
A row-count increase is a signal to investigate, not automatic proof of an error. Conversely, an unchanged count does not prove that every match is right. Check several recognisable records and any unmatched cases.
Keep order identity separate from line identity. Otherwise an attempt to remove “duplicates” may discard legitimate detail that the report needs.
Investigate the matching keys
Inspect the lookup for repeated keys and the source for missing or inconsistent identifiers. Agree whether a code should be unique within the relevant scope. If a date, location or version is part of that scope, include the appropriate business rule in the plan.
Do not delete repeated records merely to make the total return to its previous value. First establish which records are valid and how the report should represent them.
Where a match remains uncertain, keep it on a review list. That preserves the question for the person who can resolve it instead of hiding it inside a convenient filter.
Build a tiny join exercise
Create fictional source rows for one unique match, one missing match and one key with two lookup records. Write the expected result before running the merge and expansion in a test copy.
Compare actual records with that expectation. Can you explain every row that appears? Does the total reflect the intended business unit? Record the decision so a colleague can follow it next month.
The useful result is a combined table whose structure you understand. A larger number deserves a business explanation only after you have ruled out a change in how the data was combined.
Keep the reporting method connected
Explore the Make the Imported Detail Worth Summarising guide and the related practical questions below.
- Where Did the Zeros Go? When an Identifier Starts Looking Like a Number
- The Total Matches. Why Are Some Orders Still Missing?
If you want a written practice route, review the Excel Monthly Reports with Matze course overview. It requires compatible Excel Desktop for Windows with Power Query; you build the report yourself.
Check the course contents and current price. Affiliate link: Superstore may earn a commission if you buy. Check current contents, requirements and terms before purchasing.
Technical reference: Microsoft: Merge queries overview. Checked 17 November 2026. The fictional business exercise and reporting recommendations above are Superstore editorial guidance.




Discussion 1