Skip to main content

​
Why Your Data Might Be Wrong

If a report shows numbers you do not expect, or two people see different results from the same report, the cause is almost always one of a handful of things. Work through this checklist before assuming there is a bug.
This page is part of the Report Designer guide. For formula-specific issues, see the Calculated Field Cookbook.

​
The five usual suspects

1

Snapshots: are you looking at the right project version?

A report runs against one or more project snapshots (saved versions of a project’s cost data). If the wrong snapshot is selected, or a newer snapshot exists, the numbers will not match what you expect. Confirm which projects and snapshots the report is using.
2

Filters: is something narrowing the data?

Filters reduce what the report shows and they affect totals. Check both:
  • Report filters at the top (WBS, Work Package, CAM, Cost Class, Resource Type, and more).
  • In-grid filters on grid reports (header filter, filter row, search box).
A common surprise is the default Resource Type filter on charts (Labor, Non-Labor, Travel). If a category is missing, check it.
3

Permissions: are you allowed to see all the rows?

What you see is limited by your Work Package permissions. Two users can open the same report and see different rows. This is by design, not a bug. See Permissions and what each user sees below.
4

Views: is a saved view changing the layout?

A saved view changes which columns show, the order, grouping, and sorting. It does not change the data, but it can hide columns or group things in a way that makes totals look different. Switch to the default view to compare.
5

Caching: could the data be a little stale?

Report data is cached for up to 30 minutes. If a project or your permissions changed very recently, you might see the previous result until the cache refreshes. See Caching and stale data.

​
Permissions and what each user sees

For non-administrator users, the report only returns rows the user is permitted to see, based on Work Package permissions. There are three levels of grant:
GrantEffect
All (wildcard)The user sees every work package in the selected projects
WBS branchThe user sees every work package under a given WBS branch
Specific work packageThe user sees only that exact WBS and work package
Permissions also have to apply to reporting (the grant must cover the report viewer). Administrators see everything, with no permission filter.
If a colleague sees rows you do not, the difference is usually permissions. Ask an administrator to review your Work Package permissions for the project. After a permissions change, allow a few minutes for cached data to refresh.

​
Caching and stale data

To keep reports fast, Dash360 caches results for up to 30 minutes per combination of projects, fields, and filters.
  • Re-running the same report with the same filters is fast because it uses the cache.
  • Changing filters, fields, or projects runs a fresh query.
  • Cached data is automatically cleared when the underlying project snapshot changes or when your permissions change, so you do not normally need to do anything.
If you suspect stale data, change a filter and change it back, or wait for the cache window to pass, then re-run.

​
Performance: make reports responsive

Grid reports (Data Grid and Pivot Grid) run on the server: the server does the filtering, sorting, grouping, and totals, then sends back only what the screen needs. The biggest lever you control is how many rows the report has to process.
Filter early, explore later. Apply project, WBS, Work Package, and Cost Class filters before you sort, group, or expand a large report. A report scoped to one WBS branch responds far faster than the whole project.
More ways to keep reports quick:
  • Select only the projects you need. Each extra snapshot adds rows.
  • Add only the fields and columns you need. Fewer columns means less data per row.
  • Expand groups on demand instead of expanding everything at once on a huge dataset.
  • Lean on the cache. Repeating the same filtered report is fast; wildly different filters each time are not.
  • Keep cross-tabs sensible. In a Pivot Grid, avoid putting a very high-cardinality field (like many months across many years) on Columns without a time filter.

​
General pitfalls to avoid

  • Blended Cost Class. Summing Item Cost without separating Cost Class adds Budget, Actual, and Earned together. Filter Cost Class or split it with calculated fields (see the Budget vs Actual recipe).
  • Assuming views save filters. Saved views remember layout (columns, order, grouping, sorting), not filters. Filters reset on reload.
  • Renaming captions and expecting the data to change. Captions are display-only.
  • Too many categories on a chart. Filter, group small values, or pick a higher WBS level.
  • Reading an aggregate as if it were summed from rows. Aggregate calculated fields are recomputed at each group and total, so an average of averages is not the overall average. See the Calculated Field Cookbook.

​
Still wrong after all that?

If the numbers are still off after checking snapshots, filters, permissions, views, and the cache, and after confirming any calculated fields against the cookbook, capture the report name, the selected projects, the filters in effect, and what you expected versus what you saw, then contact your administrator.

​
Where to go next

Report Viewer

Run, filter, save views, and export.

Calculated Field Cookbook

Ready-to-use formulas and fixes for wrong results.