Default expanded or collapsed should follow how often the content is needed
Aliases: default expanded · accordion default · progressive disclosure rate
What it is
That a block can collapse does not mean it should start closed. Disclosure default by hit rate looks at how often a task actually uses the block: if most tasks fill it or read it, start expanded and leave existence in the page model; if only a few people, a few times, need it, start collapsed and buy space. The default is a bet on the next task’s path, not a value judgement about whether the block is “important.”
Why it happens
Starting expanded spends scroll length and buys zero discovery cost. Starting collapsed spends discovery cost and buys a shorter page. Which side pays depends on how often the block is used. At high hit rate almost everyone will open it, so collapse adds a mandatory click for all and still risks a fraction never opening; scroll would have been cheaper. At low hit rate, keeping it visible makes most people pass unused fields every time, lengthens the scroll, and interrupts the main task with a secondary block. Hit rate is not “we think it matters.” It is measured: the share of real task flows in which the block is viewed, edited, or used as a decision input. The same block can hit high for first-time setup and low on return visits; the default should fork with the scenario, not become one site-wide preference.
Studying it
In logs or tests, mark each block’s open rate, submit rate of fields inside it, and the rate of open-then-close with no edit (false opens). Contrast default-open with default-closed on main-task time, omission, and scroll depth. Independent variables: default state, heading quality, task type (first configuration / return edit). Dependent variables: effective hits on the block, time on the main path. A high false-open rate means the heading is faking existence; a high open rate with a collapsed default means the default bet the wrong path.
Where it stops holding
Content that must be seen for compliance has no vote from hit rate; it stays visible. Hit rate drifts (a cold new feature, a spike during a campaign), so the default must be revisable; do not freeze week-one cold data as permanent collapse. Personalised defaults (this person’s history) go wrong on shared devices and new accounts; fall back to the group rate. A block that mixes high- and low-hit fields should follow the high part, or be split, not averaged.
Applying it
- Count, on real tasks, how often each block is used. High hit: default open. Low hit: default closed. Write down the counting window.
- If first-time setup and return edits have different rates, give them two defaults; do not share one “user preference.”
- For a default-closed block, put the hit condition in the heading (“Expand when you need invoice fields”) so the people who should come, come.
- How to check: against main-task recordings, count how many people must open a block to finish. High share and default-closed: change the default. Low share and default-open: change it back.