Use voice command when the user speaks a bounded command phrase or dictation instruction that the product recognizes, confirms, disambiguates, and executes. Use keyboard shortcut for key chords, command palette for searchable command browsing, prompt box for AI request composition, text input for ordinary typed values, and touch gesture for pointer or touch movement.
Use voice command when the user speaks a bounded command phrase or dictation instruction that the product recognizes, confirms, disambiguates, and executes. Use keyboard shortcut for key chords, command palette for searchable command browsing, prompt box for AI request composition, text input for ordinary typed values, and touch gesture for pointer or touch movement.
Practice the judgment: Imagine this situation: Visible buttons say Send report and Export data but the speech recognizer expects hidden command names such as submit incident or download CSV. Write the person's goal, choose one of these patterns, and show the state that helps the person continue. Use the decision rules and source trail below to explain your choice.
Review the underlying principle →| Dimension | Voice command | Keyboard shortcut | Command palette | Prompt box | Text input | Touch gesture |
|---|---|---|---|---|---|---|
| UI or UX | UI + UX - Spoken command interface that listens, recognizes, confirms, and executes bounded user intents | UI + UX - Direct key combination that activates a known command, focus target, or mode | UI + UX - Modal command surface | UI + UX - Primary editable input surface for composing and submitting an AI request | UI + UX - Single-line freeform data-entry control | UI + UX - Touch-first gesture vocabulary and fallback contract for taps, swipes, drags, pinches, and multi-touch interactions |
| UI guidance | Render voice command as an explicit listening surface with microphone permission state, wake or push-to-talk trigger, listening indicator, timeout, recognized phrase, confidence, matched command, target object, alternatives, and cancel path. | Render keyboard shortcuts as discoverable command hints near the visible command, in shortcut help, in menus or toolbars when space allows, and in platform-specific form such as Command+S on macOS and Control+S on Windows. | Render a compact dialog-like command surface with a search input, current scope, typed command mode, active result, command metadata, and empty state. | Render the prompt box as a labelled, editable composer with visible draft area, send control, context chips, attachment controls, model or mode indicator when relevant, character or token boundary feedback, and clear disabled or blocked-send reasons. | Render a persistent label, appropriately sized single-line input, optional hint, visible focus state, and nearby error text that is programmatically associated with the field. | Render touch gesture affordances with visible targets, enough spacing, state feedback, and non-gesture controls for the same outcome when a gesture is path-based, multipoint, hidden, or easy to misfire. |
| UX guidance | Use voice command when users benefit from hands-free spoken control, accessibility voice input, or quick command activation and the product can manage recognition uncertainty, privacy, confirmation, recovery, and fallback paths. | Use keyboard shortcuts as accelerators for frequent, reversible, well-understood commands that already have visible controls or reachable command surfaces. | Accelerate expert navigation and repeated actions across a large product while preserving ordinary navigation for novice and low-frequency users. | Use a prompt box when users need to author a natural-language request for AI generation, transformation, analysis, or automation and must remain in control of the exact request being submitted. | Let users edit normally, validate after interaction or attempted submit, preserve their typed value after errors, and explain exactly how to fix invalid or missing text. | Use touch gesture when the product needs direct manipulation on a touchscreen, but treat gestures as part of a larger interaction contract rather than the only way to act. |
| Good UI | A mobile reporting app shows Press and say a command, records the transcript 'send report', matches it to Send report, reads back Incident 482, and requires Confirm send before uploading. | A document editor Save button shows Command+S on macOS and Control+S on Windows, disables the shortcut while a blocking validation dialog is open, and announces Saved draft. | Centered command surface with input, shortcut hint, scope chip, grouped commands, command type labels, and a visible active row. | A prompt box expands for a multi-line request, shows Shift+Enter for newline, Enter to send only when configured, and keeps the draft after a failed model request. | A reference-name field has a visible label, hint, medium width, focus ring, and error message beneath the hint when submitted empty. | A photo viewer supports pinch to zoom, double tap to zoom, plus and minus buttons, a reset button, zoom percent feedback, and a clear pan boundary after zoom. |
| Bad UI | A global voice listener treats dictated prose as commands and navigates away from an unsaved form. | A global D key deletes records while users type in comments or dictate with speech input. | Huge branded modal that buries the input below decorative content. | A blank AI field shows only Ask me anything and sends vague requests with hidden page context. | The only instruction is placeholder text that disappears when the user starts typing. | A hidden edge swipe deletes a message while the user is trying to scroll the list. |
| Good UX | A user says 'filter critical alerts', sees the recognized phrase and target list, corrects an alternative before execution, and can undo the applied filter. | A user presses Command+S in an editor, sees Saving then Saved, focus remains in the editor, and the same Save command remains available in the toolbar and menu. | Keyboard shortcut or visible trigger opens the palette, focus lands in the command input, arrows move the active row, and Enter activates the highlighted safe command. | A user selects a policy paragraph, writes Summarize risks for a non-lawyer, sees Selected text and Output: bullets chips, submits, then edits and resends the exact prompt after the first answer is too broad. | A user leaves the required field blank, sees an error only after submit, returns to the same field, and their later typed value clears the error. | A user starts swiping a gallery card, sees the next-card preview, changes their mind by moving back under the threshold, and the card snaps back with position unchanged. |
| Bad UX | A user says 'send it later' while dictating a note and the app immediately sends a message. | A speech input user says a sentence containing g and d, triggering global shortcuts that navigate away and delete a draft. | Palette is the only way to reach important navigation. | A user presses Enter expecting a new line and accidentally sends an unfinished prompt to an external model. | Validation fires on the first focus event and blocks typing before the user has had a chance to answer. | A user tries to scroll a feed but the product interprets a short diagonal movement as a destructive swipe action. |
| Best fit | Users need hands-free control, accessibility speech input, or rapid spoken command activation. | Users repeat a known command often enough that memorized keyboard acceleration saves meaningful time. | Users need to traverse a broad product surface quickly. | Users must write or revise an AI request before generation, analysis, transformation, or automation begins. | The answer is short and user-authored. | A touchscreen interaction needs deliberate design for gesture vocabulary, thresholds, feedback, target sizing, cancellation, and equivalent controls. |
| Avoid when | The spoken input is open-ended AI composition rather than a bounded command. | The command is rare, novice-oriented, high-risk, or hard to explain before execution. | The app has only a few obvious actions. | The task is better expressed as a fixed form, button, or command with known parameters. | The answer is a paragraph, comment, address block, or explanation requiring multiple lines. | The gesture is only an implementation detail inside a more specific pattern such as swipe action, pull to refresh, long press, drag and drop, bottom sheet, carousel, or map view. |
| Required state | Voice unavailable or unsupported state with non-voice alternatives. | Visible command with shortcut hint. | Closed state with discoverable trigger. | Typing state with expandable multi-line draft, send affordance, newline behavior, and character or token boundary feedback. | Empty untouched state with persistent label and optional hint. | Touch down or gesture start state that identifies the active target. |
| Accessibility burden | Keep visible labels and accessible names aligned for speech-input activation. | Make all shortcut outcomes available through visible controls or reachable command surfaces. | Use dialog semantics with a clear name and modal behavior when the rest of the page is inert. | Provide a programmatic label for the prompt editor and named controls for send, attach, remove context, clear, retry, and cancel. | Associate every text input with a visible label or equivalent accessible name that matches the visible question. | Provide a simple single-pointer alternative for multipoint or path-based gestures unless the gesture is essential. |
| Common misuse | Executing commands from partial recognition or low-confidence results. | Making a shortcut the only path to a primary command. | Hiding basic navigation behind a keyboard-only palette. | Showing Ask anything as the only instruction while hiding what sources and tools the model can use. | Using placeholder text as the only label or instruction. | Making a hidden gesture the only way to find navigation, delete, refresh, zoom, or edit controls. |