# Design for Repetition

Evaluate a workflow across repeated use, while preserving what someone needs to learn it and perform it safely. Familiarity changes the experience; it does not make everyone an expert.

## Establish the pattern of use

Infer the audience, task, frequency, consequence, and conditions of interruption from the request and available evidence. Identify assumptions when usage data or research is absent. Repetition can mean several operations in one session or returning after a long gap; those require different support.

Follow the named task once before proposing shortcuts. Then inspect what repeats: setup, navigation, decisions, data entry, waiting, feedback, and recovery. Preserve the user's scope. Reviews produce recommendations; requests to improve or implement authorise bounded changes.

## Find the accumulating cost

| Recurring burden | Useful direction | Preserve |
| --- | --- | --- |
| Re-entering stable context | Suitable defaults or retained settings | Visible current context and a way to change it |
| Repeated navigation | A discoverable direct path or return position | Orientation and a usable ordinary path |
| Repeated item operations | Batch action when items share valid semantics | Item scope, partial-failure reporting, and recovery |
| Waiting for nonessential effects | Earlier access to the completed result | Truthful completion feedback and relevant explanation |
| Repeated consequential decisions | Remember safe choices or offer a suitable preset | Review where context, cost, or reversibility warrants it |
| Interrupted work | Retain the useful draft or place in the task | An account of what was saved and what remains pending |

Measure task effort, not clicks alone. An extra comparison step may reduce costly mistakes. A shorter path may demand more memory or hide the current scope. Treat timing and error observations as evidence from the sampled conditions, not universal usability results.

## Support familiarity without requiring it

Keep the primary route understandable. Add keyboard shortcuts, repeat actions, presets, or batch operations only when they solve a supported recurring need. Make faster paths discoverable and avoid moving essential actions exclusively into shortcuts or hover states.

Retain context at the appropriate boundary. Decide whether a setting belongs to a task, session, project, device, or account. A remembered destination or filter can become stale; expose it before an action when a mistaken context would matter. Do not persist sensitive or destructive choices merely to remove a step.

Distinguish harmless acknowledgement from consequential confirmation. Simplify repetitive feedback when it obstructs use, while retaining confirmation, undo, or review where the action warrants it. Never remove an established approval requirement to make a workflow faster.

Evaluate motion and celebration over repeated encounters. Useful feedback should remain legible and interruptible where appropriate. Do not assume every expressive detail becomes annoying; inspect whether it delays work, competes with attention, or still carries meaning.

## Try the repeated and interrupted paths

Use representative sequences rather than a prescribed number of loops. Check first use, a repeat with context retained, a changed context, and an interrupted return when these apply. For batch changes, inspect mixed eligibility and partial success if the real workflow supports them.

For example, a person exporting several project revisions may benefit from retained format settings and immediate access to each finished file. Verify that changing projects does not silently retain the wrong destination. This is a hypothetical scenario, not a reproduced bug.

In a working implementation, exercise changed paths with relevant input methods and realistic data. A repeated automated sequence can expose mechanical friction; it does not establish that people feel less fatigue. Seek user evidence when the decision depends on comfort or learning. Disclose checks that remain unavailable.

For any reported bug, include the affected consumer and task, prerequisites and steps, expected and actual behaviour, and impact. Separate an observed failure from a hypothesis about cumulative frustration.

## Deliver

Explain the recurring burden, the bounded change, what the ordinary path retains, and the evidence or experiment supporting the recommendation. Implement scoped improvements when requested. Stop when the identified repetition cost is addressed without hiding necessary context, weakening recovery, or expanding the workflow into a redesign.