Too many optional fields should be deleted, not marked
Aliases: optional field bloat · mark is not enough · remove optional fields
What it is
Once optional fields fill a screen, even a clear “optional” mark cannot take them out of the effort estimate. People see the number of boxes, not the semantics of the mark. Deletion means removing those items from the main path: drop unused ones for good; move occasional ones into an expander, a settings page, or later completion—rather than leaving them on the main form as skippable weight. This entry is about the moment when marking can no longer save density. It is not about marking the minority, or writing a legend for asterisks, and not about costing abandonment field by field—that is a different line of reduction.
Why it happens
A mark changes classification, not area. Optional boxes still occupy the scan and still enter “how much longer.” Many people default to “if I can see it, I should fill it.” The more diligently items are marked, the more the main path looks like an untrimmed list. A second layer is social pressure on optional. On hiring, medical, and government forms, optional is read as “serious people fill this”; leaving it blank feels like an attitude, so the mark becomes a pressure source rather than a license. When there are enough optional items that each one needs a skip decision, the decision costs more than the input. Deletion changes the question from “leave this blank?” to “this path does not contain this item.” An expander still counts as area, but a collapsed default lowers the first-screen estimate. If the expander control is more salient than the fields, the area returns.
Studying it
Take the same optional set in three treatments: left on the main path and marked “optional,” collapsed by default under “additional information,” and removed from the flow except in settings. Instructions must say optional items may be left blank.
Independent variables: presentation of optional items (main path / collapsed / deleted), count of optional items, whether the setting is high-evaluation (hiring, application). Dependent variables: main-path completion, actual fill rate on optional items, completion time, subjective workload, judgment of whether blanks cost points.
Without real evaluation pressure in the lab, “blanks cost points” will not appear. Applications, onboarding, and expense claims are closer. A drop in first-screen bounce after collapsing is not the effect of deletion—collapsing hides, the items remain.
Where it stops holding
Fields that must remain available for compliance (emergency contact, accessibility needs) cannot vanish from the product, but they can leave the main path, closed with “you can add this now or later” before submit. Marketing wants—interests, seniority, company size—often have no compliance standing and most often masquerade as optional on the main form. On internal tools, optional fields may be shortcuts for experts; deleting them forces a detour through settings. Collapse rather than delete there. When required fields are already many, a long optional tail cannot be saved by a cleverer marking system; first-screen length is already lost.
Applying it
- Find a downstream consumer for every optional field; if no report, no business rule, and no future visit by the user, delete it from the main form.
- If it has a consumer but is not needed to finish the current task, take it off the main path: collapsed “additional information,” post-success completion, or account settings—do not keep it bound beside required items with an “optional” tag.
- Name the expander with how many items are inside and that they are unrelated to the current task; avoid empty labels like “more information” that reopen the area.
- Verify by removing these items from the main path and comparing completion rate and time in the same window. Completion up and later-completion usage near zero means those optional fields were never needed, not that marking rescued them. If fill rate on optional items still tracks required in a high-evaluation setting, move them off and recheck whether “will this cost me points” falls.