Framework · Enterprise UX · 3 min read
Expert-user efficiency
Seven design patterns for internal tools used by experts doing repetitive work (efficiency features, density, consistency, progressive disclosure, error prevention, respect for muscle memory, strong data tables) that you can use to review and improve a screen.
Keyboard shortcuts, bulk actions, saved filters, defaults, inline editing.
e.g. change 40 statuses at onceExperts want more data on screen. Don't add whitespace for its own sake.
e.g. compact table rowsOperators switch apps all day; differences cost time and cause errors.
e.g. one filter pattern everywhereHide rare, complex features until the user needs them.
e.g. customs fields only when neededConstraints, validation, confirmation for destructive or bulk actions, undo.
e.g. confirm before bulk deleteBig redesigns break learned behaviour. Change incrementally and tell users.
e.g. keep old shortcuts workingThe core UI: sorting, filtering, column config, export, row actions.
e.g. start every review hereUse it when
- A redesign proposal adds whitespace, larger cards or a guided tour to a screen that operators use all day.
- Operators complain that a new version is "slower", even though it looks cleaner.
- You review a core screen, such as a data table or an entry form, and want a checklist for expert use.
How to use it
- Pick one high-frequency task. Choose a task that experts do many times a day, such as creating or correcting a booking. Watch someone do it (see Contextual inquiry). Output: one task with its steps and how often it happens.
- Count the friction. Note every mouse trip, page change, repeated entry and wait. Note the workarounds too. Output: a list of friction points for that task.
- Check it against the seven patterns. For each friction point, ask which pattern would help: efficiency features (keyboard shortcuts, bulk actions, saved filters, defaults, inline editing), density, consistency across tools, progressive disclosure, error prevention, respect for learned behaviour, data tables. Output: a friction point mapped to a pattern.
- Start with the data table. Data tables are the core UI of most internal tools. Check sorting, filtering, column configuration, export and row actions first. Output: a short list of table fixes.
- Change incrementally and say so. Big redesigns break muscle memory. Ship small changes, tell users what moved and why, and keep old shortcuts working where you can. Output: a release plan in small steps with a change note.
- Measure the task. Compare task time and errors before and after. Output: a before/after number for the task.
Worked example
An internal operations platform with 6 product teams. A team reviews the "correct a booking" task, which operators do many times per shift.
| Friction seen | Pattern | Change |
|---|---|---|
| Operators open each booking to change one field. | Efficiency features | Inline editing in the table. |
| They change the same status on many rows, one by one. | Efficiency features | Bulk action with a confirmation step. |
| They filter by the same three values every morning. | Efficiency features | Saved filters. |
| A delete button sits next to save. | Error prevention | Move it, add confirmation and undo. |
| Rare customs fields fill half the form. | Progressive disclosure | Show them only when the shipment needs them. |
| The table filter works differently from the one in the billing tool. | Consistency across tools | Use the shared table pattern from the design system. |
The team shipped these as three small releases, with a short note to operators each time.
Watch out for
- Consumer habits. More whitespace, fewer columns and onboarding tours slow experts down. Density is a feature here.
- Big-bang redesigns. A full new layout breaks the speed experts built over years. Change in steps and communicate each one.
- "They have no choice anyway." Users with no choice build workarounds, make errors and need more training. That costs money.
- Bulk actions without safety. Bulk and destructive actions need confirmation and, where possible, undo.
Template
Copy this checklist to review one expert screen.
Expert screen review
Screen: Task: Uses per day:
[ ] Efficiency: shortcuts, bulk actions, saved filters, defaults, inline edit
[ ] Density: the data experts need is on screen without scrolling
[ ] Consistency: same patterns as the other tools they use
[ ] Progressive disclosure: rare, complex features are hidden until needed
[ ] Error prevention: constraints, validation, confirmation, undo
[ ] Learned behaviour: nothing moved without a reason and a change note
[ ] Data table: sort, filter, column config, export, row actions
Top 3 changes:
Task time / errors before: after: