The gulf appears as not knowing what to do

What it is

The typical sign of a Gulf of Execution is users remaining at the goal level: scanning the screen, hovering over icons, right-clicking for menus, or asking "where is the setting?" instead of performing a clear action. It appears as a missing, invisible, disabled, terminologically mismatched, or command-only entry. This card follows directly from "the gulf is the distance between intention and executable operation": that card gave the definition and the way to measure it; this one describes what it actually looks like in real use — not an abstract distance figure, but a string of concrete, observable stuck behaviors.

Why it happens

Whether a user can start an action depends on whether they can form a candidate list of actions in their head, and the only raw material for that list is whatever cues are currently visible on screen — this is the key point: users do not reason abstractly about what the system supports, they guess "what can I do with this" from what they see in front of them. If the interface offers only abstract icons, functions discoverable only by right-clicking, gestures triggered only by a long press, or a plain command syntax, with no readable entry sitting in the open, the candidate list is empty or wrong, and the user's only lifeline is habits carried over from other software — try a right-click, try a long press. This is exactly where "trial and error" comes from; it is not impatience, it is the interface failing to supply enough cues for the candidate list to form normally. A subtler case exists too: the system genuinely supports the goal, but that support is hidden — "supported" and "discoverable" are two separate things, and a disabled control that does not explain its condition tells the user not "you can't do this yet, but you can once a condition is met" but "this simply cannot be done," misclassifying a genuinely feasible task as infeasible in the user's mind.

Where it stops holding

Not every function needs to stay permanently visible on screen. Low-frequency advanced functions can reasonably be tucked away with progressive disclosure, provided that a discoverable entry remains and its label lets the content be predicted — a label like "more settings" works precisely because, while it does not name the specific function, it clearly tells the user "there is more this way," which is a different thing entirely from hiding something with no cue at all. Nor should every brief hesitation from a user be read as evidence of a gulf: a skilled user pausing mid-task is often just thinking about which option to pick next, ordinary cognitive processing time, not an inability to find the entry. Distinguishing the two requires combining the user's current goal, the task stage they are in, and what they do right after the pause — a pause followed by aimlessly scanning the whole interface is a gulf; a pause followed by clicking directly on the expected option, just taking a moment to decide, is not.

Applying it

• Observe the first minute after a user states their goal, and log which icons they hover over, whether they try a right-click, whether they repeatedly go back to the previous page, or whether they turn to a search box — this combination of behaviors is the most direct behavioral evidence that "the candidate list is empty."
• Give a previously hidden action a visible entry, a short explanation in a menu, or make it searchable; for a disabled control, state the specific condition for it being disabled, rather than letting it look like a function that never existed.
• An empty-state screen should list what a user can specifically do here, not just display "no data yet" — an empty state is often the first moment a user encounters a given feature area, which is exactly when an action candidate list is needed most.
• How to check: use a "give the goal, no hints" test to measure whether a user can find and open the correct entry within a reasonable time on their own; track this rate as a quantifiable metric rather than stopping at a subjective impression like "users say this is hard to find."

Related

• Same group: B4.02.1 The gulf is the distance between user intention and executable system operations · B4.02.3 Visibility, signifiers, and constraints narrow it
• Nearby: B2.09 Discoverability · B3.06 Recognition over Recall
• Search terms: discoverability · hidden action · trial and error · action candidate