Test the same reporting task

Video transcript

1. 0:00–0:06

   Begin with one report • Agree the record grain and measure before features. • One order · agreed formula · fixed snapshot • Power BI • Data Studio • Retain any extra preparation each trial needs

   One agreed record grain, formula and fixed source snapshot branch to equally sized Power BI and Data Studio candidates. Extra preparation required by either trial stays explicit. No equality of effort, totals or product quality is claimed.

2. 0:06–0:12

   Trace the calculation • Connected tables do not prove data integrity. • Order A · branch B2 • Branch lookup: B1 only • B2 has no matching branch • Also test repeated rows and filters

   In a hypothetical example, order A refers to branch B2. The lookup contains only B1, so a traced join exposes an unmatched B2 key. A relationship does not repair source integrity. Repeated rows and filters also need testing; no real data-quality result is asserted.

3. 0:12–0:18

   Test the real audience • Confirm roles, credentials and distribution rights. • Viewer role • Permitted export • Actual viewer: agreed rows only? • Export: agreed rows only? • Developer access is not the viewer test

   The actual viewer role and export permission crops lead independently to questions about the agreed visible and exported rows. Both checks stay open. A developer account does not stand in for the intended viewer, credential mode or distribution arrangement.

4. 0:18–0:24

   Observe a known change • Refresh controls do not remove upstream delays. • Source changed • Preparation + cache • Report observed • V2 • Measure delay · separately test a failed connection

   A hypothetical change labelled V2 moves from source through preparation and cache to the observed report. The animation demonstrates stages, not measured or promised refresh timing. The real trial must measure delay and separately exercise a failed connection.

5. 0:24–0:30

   Choose from trial evidence • Include preparation, operation and ownership. • Power BI • Data Studio • Trial evidence + named maintainer • Include preparation, access and operating costs · choice open

   The two equal candidate reports lead to an open trial-evidence record with a named maintainer. Include preparation, access and operating costs alongside reconciliation, language, update and failure tests. No winner or acceptance outcome is shown.