Signifier

What it is

Norman's distinction: an affordance is what an object can be used for (a button can be clicked); a signifier is the visible cue that tells the user it can be (the raised appearance, the pointer cursor on hover, the "Submit" label). Affordances exist in the world; signifiers are what designers actually control.

In digital UI, signifiers appear in visual weight (buttons that look pressable), state (hover, focus, active), cursors, icons, text labels, and animated hints. The pointer cursor turning into a hand on a hyperlink is a signifier. The chevron suggesting a dropdown is a signifier. The three-dot menu icon is a signifier (and, debatably, a weak one).

Why it matters

Interfaces with weak signifiers force users to hunt — hover to find what's clickable, tap to discover what's interactive, read the whole page to find the primary action. Interfaces with strong signifiers let users act with confidence. On touch, where hover doesn't exist, signifiers are the entire pre-tap signal.

When to use it

Every interactive element. Primary CTAs, secondary actions, destructive controls, drag handles, expandable sections, dismissible overlays, form inputs, links in body copy. Any element that can be acted on needs a signifier that tells the user both that it can be acted on and what the result will be.

What it looks like in practice

A primary button with filled background, consistent border radius, pointer cursor on hover, and a verb-first label ("Save changes" rather than "OK") carries clear signifiers. A "ghost" button with only a thin border and no icon reads as decorative to many users — the signifier is weak.

A card that is clickable needs more than just hover state — raised appearance, pointer cursor, and ideally a chevron or arrow icon signalling navigation. Cards that look identical to non-interactive cards fail the signifier test on first visit.

What people get wrong

• Removing signifiers for minimalism. Ghost buttons, invisible drag handles, hover-only states and "flat" everything strip the signifier from the affordance. The affordance still exists; the user just can't see it.
• Inconsistent signifiers. If some clickable cards have hover elevation and others don't, users learn not to trust the signifier at all.
• False signifiers. A box styled to look like a button that isn't one erodes trust in buttons across the whole interface.
• Signifiers that depend on motion. If the user has prefers-reduced-motion, animated hints disappear; design the signifier so it works without the motion.

Learn next

Signifier is the operational cousin of affordance — one names what's possible, the other names what the user can see. Both relate directly to mental model (does the signifier match what the user expects to happen) and to gestalt principles (similarity creates signifier equivalence).

Reviewed 1 October 2026 · Editorial standards