Old habits carrying over after a redesign

What it is

After an interface is redesigned, the action sequences users trained on the old version don't vanish just because the interface changed — they're still there, and they're often re-triggered by similar-looking cues in the new interface, so what the user actually does is the old version's action, not what the new interface actually calls for. This is capture error showing up in a redesign context: what hijacks the user this time isn't another path coexisting inside the current interface, but an old path that's supposed to no longer exist but is still the most familiar one in memory.

Why it happens

This variant of capture differs slightly from the coexisting-paths case: the competing "path" doesn't actually exist in the current interface anymore. The user is triggered by old cues the new interface happens to have retained — a roughly similar position, a similar appearance, a similar timing — which reawaken the old action. The more superficial similarity the new interface retains, the more likely the old action gets triggered; and once it's triggered and carried out, the outcome can be a dead click (the old target no longer exists), an error, or something worse — the old target's location has now been repurposed for a different function, and the old habit precisely fires a new function the user never meant to use at all.

Where it stops holding

This effect doesn't fire for every redesign — it has a non-monotonic relationship with the magnitude of change. If a redesign is large enough, the cues that used to trigger the old action disappear altogether, and users shift into "relearning" mode rather than "erroneous execution" mode, so carryover is less visible. The worst range is precisely the moderate change — "looks about the same, but the details changed" — where the old cues survive intact while the meaning or outcome of the action has actually shifted.

Applying it

When evaluating a redesign proposal, specifically measure how similar the key trigger cues are before and after — position, color, and timing in particular, since these are the cues most likely to trigger an action below the level of deliberate attention — and check whether any of them were kept while their meaning changed (the same-position button now does something else). This "same place, different meaning" pattern is where carryover errors concentrate, and it should be flagged during design review rather than discovered after launch. Verification: in the first few weeks after a redesign ships, specifically track whether mis-operations at the old trigger locations are significantly higher than elsewhere in the interface. If they are, that confirms a genuine carryover problem rather than a generic learning-curve effect.

Related

• Same group: A10.05.1 A familiar sequence hijacks the current intention · A10.05.2 A longer shared prefix and a bigger frequency gap amplify capture · A10.05.4 Post-change transition protection instead of a one-shot cutover
• Nearby: A6.09 Procedural memory and automatization
• Search terms: habit carryover · interface migration error · post-redesign slip