On this page11 sections
Growth Operations ResearchA propagation test across sharing paths, not a compliance audit.

Testing Whether Dashboard Access Changes Reach Every Shared Link

A bounded test for checking whether changing dashboard access actually updates every shared link, embed, and scheduled subscription.

Published ••5 min read
Testing Whether Dashboard Access Changes Reach Every Shared Link research illustration

Why sharing paths matter

Dashboard access is rarely controlled in one place. A report may be shared through a direct link, an embedded view, a scheduled email, a team folder, or an export that someone saved. Changing the access setting on the dashboard does not automatically update every path that already reaches the data. This test checks whether a deliberate access change actually propagates to the shared links and subscriptions that carry the dashboard outside its original audience. It is a propagation test, not a judgment about who should have access.

Define the access change under test

Start by naming the change and its intended effect. It might be removing a person from a dashboard, changing a link from public to restricted, or moving a report into a folder with different permissions. Record the change date, the person who made it, and the expected outcome. If several changes happened together, split them where possible so the result is attributable. A test that bundles a permission change, a folder move, and a new link type cannot tell the team which action failed.

Inventory every sharing path

List the ways the dashboard can reach a person: direct links, embedded views, scheduled subscriptions, exported files, mobile apps, and team spaces. Record the owner of each path and when it was last used. Include links that were shared in chat or email, because those are often the ones nobody remembers. The inventory does not need to be exhaustive on the first pass, but it should cover the paths the team actually uses. A path that is not listed cannot be tested.

Capture the before state

Before the access change, test each path with a controlled account. Record whether the account can open the dashboard, what data it sees, and whether it can export. Save the exact URL, the timestamp, and the account used. This baseline is what makes the test meaningful. Do not rely on the sharing dialog's description of who has access; it may not reflect an embed or a forwarded link. The actual rendered view is the evidence.

Apply the change and recheck each path

After the change, retest every path with the same accounts and the same URLs. Record whether access was removed, retained, or partially removed, such as a dashboard that opens but shows blank tiles. A partial result is common and should be its own category. Note the time of the check and whether any cache or session refresh was needed. Immediate checks catch some paths; others may only reflect the change after a scheduled refresh or a re-login, so record the delay where it matters.

Test the scheduled subscriptions

Scheduled email subscriptions are a frequent leak because they run under the scheduler's credentials rather than the recipient's. Check whether a subscription still delivers after the access change, and whether the recipient can open the attached dashboard. Record the subscription name, the owner, the schedule, and the result. If a subscription continues after access is removed, record it as a high-priority finding. The email may contain a snapshot of the data even if the live link is blocked.

Test embeds and exported files

An embedded view on a website, wiki, or internal portal may use a token or a public link that bypasses the dashboard's user permissions. Open each embed in a private browser session and record what it shows. For exported files, check whether a previously exported file remains accessible to people who should no longer see it. The team may not be able to revoke an already-downloaded file, but it can stop future exports and record the exposure. The result should distinguish a live embed from a historical file.

Classify each path

Group the results into access removed, access retained, partial access, and untestable. Add a category for a path that was removed entirely, such as a deleted subscription. The classification shows whether the change worked as intended and where the leak remains. It also gives each finding a clear owner: the dashboard owner for permissions, the subscription owner for scheduled sends, and the platform owner for embeds. Without categories, the team sees one ambiguous result and no repair path.

Quantify the propagation gap

Report the number of sharing paths tested, the number that reflected the change, and the number that did not. Report counts alongside percentages and show the high-risk paths separately. Do not describe the result as an industry benchmark or a compliance score. It describes this set of paths at this time. A single retained subscription that emails an executive dashboard to a departed contractor deserves more attention than several stale links that no one uses.

Choose the repair that closes the path

For each retained path, decide whether to delete the link, rotate the embed token, change the subscription owner, or rebuild the sharing path under the correct permission model. Record the decision and the owner. Prefer removing a redundant path over adding another permission layer, because fewer paths mean fewer places for access to drift. Retest the same paths after the repair. A repair that closes the leak without breaking the intended audience is the goal.

Document limits and data handling

The test depends on access to every sharing path and on controlled test accounts. It cannot find links that were shared outside the systems the team can inspect, and it cannot prove that a path is unused just because no one remembers it. State those limits in the method note. Use test accounts with no real customer data, restrict access to the evidence, and delete any exported test files. A repeatable test is one that another administrator can run before the next access change, not a one-time investigation after an incident.

Sources

  1. https://www.nist.gov/cyberframework
  2. https://www.census.gov/data/developers/guidance/api-user-guide.html
  3. https://knowledge.hubspot.com/records/understand-original-source-properties
  4. https://www.w3.org/TR/NOTE-datetime
  5. https://www.ftc.gov/business-guidance/privacy-security