Power BI drill-through without losing filter context

Drill-through should help a reader inspect the records behind a selected result. Define which filters travel to the detail page and which measures or eligibility rules must remain consistent. A detail table that opens successfully can still show a different population from the summary the reader selected.

Use a controlled example to test the handoff before adding more pages. The retail modelling lab supplies small enough data to inspect every contributing line.

Define the intended journey

Start from a category visual showing Paid Net Paise by DimProduct[Category]. Select January 2026 and then Software. The selected value is 29,000 paise, contributed by S1 at 19,000 and S3 at 10,000.

Create a detail page with Category as its drill-through field and a table showing LineID, OrderID, OrderDate, CustomerKey, Status and calculated net line value. Configure the filter handoff so the January period is preserved as intended.

Microsoft's drill-through documentation describes the target-page fields and navigation options. Review the current Keep all filters setting and its effect in your actual report; do not infer context solely from the destination page title.

Preserve measure eligibility explicitly

The source measure defines Paid through DAX. That measure condition is not necessarily the same as a visible page filter that automatically restricts every raw detail row. A detail table containing all statuses can therefore diverge from a Paid-only summary unless its eligibility is designed deliberately.

Add an explicit Paid restriction to the detail population or use a detail design that reliably includes only the measure's qualifying records. Test a category with Pending activity, not only Software, where every fixture line happens to be Paid.

The South/Support pending line S5 is particularly useful: it must not appear as a contributing line for a Paid-only result. A test using only paid records would miss this defect.

Reconcile the exact key set

For January Software, the expected detail LineID set is {S1, S3}, and its net values sum to 29,000. Across all dates, Software adds S6 and totals 39,000. If the detail page shows S6 under January, the period context was lost or overridden.

Check both keys and amounts. Two different record sets can have the same total, and a summary total can remain correct while the detail includes extra zero-value rows.

Display the selected category and period in a scope note on the detail page. If several values are allowed, use a label that handles multiple selections instead of displaying a misleading single category.

Test the return journey

Provide a clear back action and verify that returning preserves the reader's original context. Test repeated navigation between different categories so stale filters do not persist unexpectedly.

Also test direct navigation to the detail page if it is visible through another route. Decide whether an unfiltered detail page is useful, should explain that a selection is needed or should present a safe default population. Do not assume every visitor arrived through the same source visual.

Bookmarks and page navigation can interact with saved filter state, so include them in the test if they are part of the report's actual journey.

Keep drill-through separate from security

A hidden detail page or a restrictive drill-through filter is not a substitute for row-level security. The consumer's allowed rows must remain restricted regardless of navigation route.

Run the same journey under the relevant consumer role. For a North-only role, January Software should expose only S1 and 19,000 paise. A visible summary and its detail should agree within the authorized population.

The lab's SQLite checks establish source arithmetic. It does not run Power BI navigation, DAX or RLS; record those application results before approving the report.

Use a small acceptance matrix

Test January Software, all-period Software, a category with Pending activity, an empty combination, a second navigation after the first and a restricted consumer. For each, record expected filters, keys, total and back-navigation state.

Exercise: deliberately disable period preservation and compare the January Software detail with its 29,000 summary. Restore the intended setting, then test whether a Paid-only summary still needs an explicit status rule on the detail table.

NeuraPath's Data Analytics with Generative AI course connects dashboard interactions with reconciliation. A good drill-through page makes the selected population more inspectable without changing its meaning.

Continue learning

This article is part of the Power BI data models and reporting sequence. Use the neighbouring tasks when you need the prerequisite or the next application.

• Review the prerequisite or neighbouring task in Build a sales dashboard around three business decisions.
• Continue with Design dashboard tooltips that explain a metric.