Dashboard issues
This guide covers the most common problems you'll encounter with Mammoth dashboards—blank screens, stale data, AI generation failures, sharing errors, and mobile display issues—and gives the fix for each.
Dashboard not loading (blank screen)
A blank dashboard can mean several things. Work through this checklist in order:
Step 1: Check platform status
Visit https://status.mammoth.io. If there is an active incident affecting dashboards, wait for the incident to resolve.
Step 2: Check View availability
Every dashboard is connected to a View as its data source. If the underlying View has been deleted, the dashboard loses its data source and will show blank.
To check: open the source project, navigate to the View the dashboard uses, and confirm it still exists.
Step 3: Check your permissions
Everyone in the dashboard's project has access to it, and people outside the project can be invited as viewers. If you were recently removed from the project or your role changed, the dashboard may no longer open for you.
Contact your workspace admin to verify your project access.
Step 4: Hard refresh the browser
Press Ctrl+Shift+R (Windows/Linux) or Cmd+Shift+R (Mac) to bypass browser cache. If the dashboard loads after this, a stale cached state was the cause, so no further action is needed.
Dashboard showing wrong or stale data
How dashboard data refresh works
A dashboard has a draft and a published version. New data from the dashboard's source Views comes into the draft when you sync, and viewers only see it once the draft is published.
To get fresh data in your dashboard:
- First ensure the underlying dataset and View are up to date. If the dataset is on a scheduled refresh (via an automation), wait for the refresh to complete, or trigger it manually.
- Open the dashboard's Sync settings and click Sync all sources, or turn on Sync automatically so new rows come into the draft whenever the View's pipeline runs.
- Publish the dashboard, or turn on Publish new data automatically so new data goes live as soon as it syncs. Layout and theme changes still wait for you to publish.
What to do if viewers see stale data: Check the Sync settings. If the draft is up to date but viewers still see old values, the dashboard has unpublished changes. Publish it.
Dashboard still shows old data after publishing
If the source View has current data and the dashboard is published, but it still shows old values, check whether the dashboard has a saved filter state that excludes current data. Open the dashboard filters and reset all filters to their defaults.
AI dashboard generation failed or produced unexpected results
Generation took too long
Generation can take a while for wide or complex Views.
Fix: Refresh the page and try again. If it keeps failing:
- Reduce the number of columns in the source View. AI generation works best with focused Views
- Break a wide View into multiple narrower Views
- Simplify the prompt. Be specific about the few charts you want rather than asking for a complete analysis
Generation produced wrong chart types or unexpected data
The AI generates charts based on column names and data types in your View. If the output doesn't match your intent:
- Refine your prompt. Be explicit: "Show me a line chart of revenue by month and a bar chart of units sold by product category" produces better results than "analyze my sales data."
- Check column naming. Columns named
col1,field_3, or other generic names confuse the AI. Rename them to meaningful names in your pipeline before creating the View. - Check data types. If a date column has been imported as text, the AI cannot create time-series charts from it. Fix the column type in your pipeline first.
Sharing problems
Can't find the share button
Click Publish & Share in the editor toolbar, or open Share from the dashboard. You need edit access to the dashboard to publish or share it.
Publishing limit reached
Mammoth limits the number of published dashboards based on your plan:
| Plan | Published dashboards |
|---|---|
| Free | 3 |
| Starter | 25 |
| Team | 100 |
| Pro | Unlimited |
When you reach the limit, Mammoth tells you that your plan's published-dashboard limit is reached. To resolve: either Unpublish an existing dashboard, or upgrade your plan.
For plan upgrade options, contact support@mammoth.io.
Shared link not working for recipients
When a shared dashboard link fails for the recipient, check:
| Access type | Common cause | Fix |
|---|---|---|
| Anyone with the link (Public) | The dashboard was unpublished | Republish it. The same link works again |
| Sign-in required (Authenticated) | Recipient isn't in the dashboard's project and wasn't invited | Invite them as a viewer, or change the access type to Anyone with the link |
Older dashboards may also use password-protected access. If the password was changed after you shared the link, send recipients the new password.
Layout and rendering problems
Charts overlapping or misaligned
If charts look uneven or misaligned:
- Open the dashboard in edit mode
- Use Tidy up to line up the tiles on the page (even widths, flush edges, one height per row), or Edit layout to drag and resize tiles
- Save when the layout looks correct
Reset to auto layout returns the page to the automatic layout.
Dashboard theme not applying correctly
If a dashboard looks visually inconsistent (fonts, colors, or spacing look wrong):
- Open the dashboard in edit mode
- Open the dashboard's design options and select the theme you want
- Save the dashboard
Mobile display issues
Tablet and mobile layouts adapt automatically from the desktop layout, so build the layout on desktop and check how it reads on a smaller screen.
Tip for mobile-critical dashboards: Place the most important charts at the top of the page, so they are the first thing a mobile viewer sees.
Dashboard loads slowly
If a dashboard takes a long time to load, the bottleneck is usually the underlying View query, not the dashboard itself.
Diagnose the View:
- Navigate to the source dataset in your project
- Open the View the dashboard uses
- Check the View's row count. Views with very large row counts are slower to query
Optimization strategies:
- Add filters to the View to reduce its row count before the dashboard queries it
- Use aggregated Views (group and summarize data in the pipeline) rather than passing raw row-level data to dashboards
- If the dashboard shows key metrics and aggregate charts, the View should contain aggregated data, not transaction-level rows
Still not resolved?
If none of the above resolved your issue, contact support@mammoth.io with:
- Your workspace name and dashboard name
- The steps you took before the issue occurred
- Any error messages displayed
- A screenshot of the problem