Skip to content

Contextual Hint or Tooltip

Interface artifact — instantiates Latent Affordance Surfacing

A small, in-context cue that names an available action at the moment and place it becomes relevant, then gets out of the way.

Contextual Hint or Tooltip is the actual cue — a transient, in-context signifier layered onto the live interface that names a latent affordance at the moment the user's task makes it relevant, and then disappears. It is not an experiment about placement and not a screen for an empty container; it is the small artifact itself — a hover tooltip, a coach-mark, an inline "did you know" nudge — whose whole design is a signifier plus a trigger. Its defining idea is just-in-time relevance: rather than teaching everything up front or adding a permanent label to a crowded surface, it waits for the context that makes one specific affordance worth knowing, delivers the cue there, and gets out of the way once the point is made. Timing and context are the design, not decoration on it.

Example

A desktop spreadsheet app has a Pivot Table feature that most users never find; they instead retype summaries by hand from large tables. A contextual hint watches for the trigger that makes the feature relevant — the user selects a wide, many-row tabular range — and, only then, shows a small coach-mark anchored to the toolbar: "Summarize this as a Pivot Table →", with the target button highlighted. It is dismissible, it remembers dismissal, and it fires once.

The affordance was always present in the toolbar; what was missing was any reason to notice it at the instant it would have helped. The cue supplies exactly that: a named signifier ("Pivot Table") placed at the moment of relevance (a big selection), pointing at the real control. A user who takes it is carried into the feature; a user who dismisses it is not nagged again — and either way the hint has done its one job and vanished.

How it works

  • Bind to a trigger. Define the context that makes the affordance relevant — a state, an action, or a struggle signal — so the cue appears because the moment warrants it, not on a timer.
  • Author the signifier. Give the cue a name and a pointer that make the target action and its location legible at a glance: what it is, and where to go.
  • Set a display-and-dismiss policy. Decide persistence, frequency cap, and dismissal memory so the cue teaches once and then yields; a hint that outstays its welcome becomes noise.

The artifact is distinguished by riding the moment of relevance: it earns attention by being contextual and temporary rather than permanent and everywhere.

Tuning parameters

  • Trigger sensitivity — how eagerly the cue fires. Too eager and it interrupts unbidden; too shy and the relevant moment passes unmarked.
  • Signifier richness — icon, short label, or a sentence with a pointer. More content teaches more but slows the glance and crowds the surface.
  • Persistence — hover-only, sticky until acknowledged, or a one-time coach-mark. Stickier cues are harder to miss and easier to resent.
  • Frequency cap and dismissal memory — how often it may reappear and whether a dismissal sticks. Weak memory turns a helpful hint into a recurring nag.
  • Targeting — show only to users who have not yet used the feature. Untargeted hints teach veterans nothing and train everyone to dismiss on sight.

When it helps, and when it misleads

Its strength is the highest relevance-to-effort ratio of any surfacing cue: it teaches an affordance inside the flow, exactly when it helps, and then removes itself, so a user learns without leaving the task. Well-targeted, it is close to free for those who already know the feature.

Its failure modes all stem from overuse. Fire too often and users develop learned dismissal — they swat away every hint reflexively, so even a genuinely useful one is gone before it registers; the honest name for a first-run wall of coach-marks is a tour nobody reads. Hints also paper over deeper discoverability problems: reaching for a tooltip when the real fix is a clearer permanent signifier on the surface treats the symptom and leaves the affordance hidden the moment the hint stops firing. And the cue is a natural home for misuse — dressing a promotion as a "helpful tip." The discipline is strict: fire only on true relevance, cap frequency, remember dismissal, and prefer fixing the base signifier[1] when a hint would merely mask its absence.

How it implements the components

Contextual Hint or Tooltip realizes the cue-authoring side of the archetype — the perceptible signifier and the moment it is delivered:

  • affordance_signifier_design — its core: the hint is the signifier, a name-plus-pointer authored to make one latent action legible at a glance.
  • surface_placement_and_timing_rule — the trigger binds the cue to the moment and place of relevance; just-in-time timing is the artifact's defining property.

The empirical question of which placement wins, the non-visual channel, and the empty-container case — salience_and_noise_calibration and discovery_measurement_loop, modality_and_accessibility_redundancy, negative_space_or_empty_state_cue — belong to Call-to-Action Placement Test, Assistive-Technology Labeling Pass, and Empty-State Prompt; this artifact designs the cue, not the experiment that positions it.

References

[1] Don Norman's signifier — a perceptible cue that tells the user where and how to act — is precisely what a contextual hint provides on demand. His point that a good signifier belongs on the surface itself is the caution here: a hint should surface an affordance the design failed to signal, not become a permanent substitute for signalling it. ↩