A report diagram shows an arrow from source to model, yet it does not explain why a particular request disappeared from the final number. Connected items are useful context; explaining a value also requires transformation, filter and record evidence. Establish which level of traceability you actually purchase with a dashboard.
Start with an identifiable number
Do not request “data traceability” without an example. Select a displayed value with its page, reading time, filters and authorized viewer context. Date, branch or access boundaries can change the result. Comparing screenshots from different contexts does not independently establish a transformation fault.
This is neither redefining every company metric nor choosing a new warehouse. Once meaning is agreed, ask which data and transformations produced this reading and what available evidence explains. A manually documented step may be needed where platform tooling lacks coverage.
A fictional excluded-request test
A venue-preparation report counts requests by branch and period. An employee says a known request is missing. This educational example is no performed investigation. Filters, transformations, merging, extraction timing or access could matter; the buyer should not select a cause before examining evidence.
Ask the supplier to trace an authorized sample from source through stages into the result, identifying request references before and after transformation. If identifiers change or records aggregate, show the connecting evidence. An arrow between items does not alone explain a particular record's fate.
A trace card for one number
Use this card before signing report scope. Requesting traceability grants no access to all source data. Agree an appropriate sample and presentation that reveal nothing beyond the viewer's authorization. Document hidden portions and inference limits rather than bypass permissions.
| Stage | Requested evidence | Gap to disclose |
|---|---|---|
| Displayed value | Explicit filters and reading time | Different comparison contexts |
| Transformation | Relevant rule or transformation-version reference | A step the platform does not automatically draw |
| Source and sample | Extraction reference and authorized records | Unavailable record or unretained history |
| Connection to result | Sample inclusion, exclusion or aggregation explanation | Unestablished portion and follow-up owner |
When transformations have different versions, retain the reference used for the reviewed reading rather than only today's version. Today's diagram cannot reconstruct an unretained historical state. Ask what the project preserves and what additional work historical traceability would require.
What does a Fabric diagram mean?
Fabric documentation describes workspace-item relationships and external sources one upstream level away, with visibility and connection-method limits. This useful dependency map is not automatic proof of every contributing row or transformation. A source hidden from a viewer does not imply its absence.
Ask what the diagram covers in the actual design, where work occurs outside the workspace and which steps need other references. Do not require a particular product's interface when different tools are used. Define the question and evidence, then let the specialist explain a suitable method and its limits.
A detail page is not necessarily raw source data
Power BI describes drillthrough as navigation to a report page filtered by context. This can explain a selected group without independently establishing original records and every transformation. Distinguish useful page navigation from provenance evidence across processing stages.
In the example, a detail page might show the correct branch's requests while leaving the excluded request unexplained. Identify the stage affecting inclusion or record that it remains unestablished. A final report should not imply complete traceability while leaving unknown portions undisclosed.
Connect this with dashboard scope and metric definitions. Send a value, its context and its trace card when requesting data services from Nextwave to specify reviewable evidence within the implementation proposal.
References and reading limits
Scope and limitations of Fabric lineage and the drillthrough concept were read on 4 November 2026. No data examination or export was performed. Prepared with AI assistance and information and translation review.
