A useful dashboard answers a small set of questions from a known dataset. Begin with the fields you can collect, then choose charts. A chart cannot create missing private metrics or recover uncollected history.
Choose the dataset
Keep three sources distinct: native analytics for your authorized account, captured public-profile data, and an imported account archive. Their fields and reporting windows differ. Record source, scope, capture date and missing values before combining them.
Tweet Archivist profile analytics use captured post counts, activity and available views. The account-archive upload is a separate workflow. The product does not offer every native analytics field or a general custom-dashboard builder.
Specify each dashboard card
| Card | Definition | Context to show |
|---|---|---|
| Posts captured | Distinct post IDs in the selected dataset | Period and collection gaps |
| Public interactions | Likes + reposts + replies | Capture time and post-type mix |
| Interaction rate | Defined actions divided by available exposure count | Denominator and excluded rows |
| Follower count | Latest recorded profile snapshot | Snapshot date, not implied live value |
| Website outcome | Visits or conversions recorded at your site | Attribution window and tracking gaps |
Select charts for decisions
Use a time series for recorded change, a ranked table to inspect posts, and a grouped comparison for topics or formats. Show sample size next to each average. Avoid a red/green “good rate” grade based on an unsourced universal benchmark.
A hypothetical dashboard with 12 tutorials and 3 announcements should not suggest the comparison is balanced. Add the underlying posts so readers can inspect the difference.
Validate and review
- Check one card against source rows.
- Confirm IDs are not rounded and duplicates are removed.
- Check empty and zero-denominator cases.
- Confirm filters apply consistently.
- Show last update and known gaps beside results.
See native access and rate calculations for the supporting tasks.