A website audit without a form is a mood. You click around, notice things, maybe jot a few notes — and two weeks later the findings have evaporated, the worst issue is unfixed, and nobody can say what the audit actually concluded. The difference between an audit that changes a site and one that just makes you anxious for an afternoon is embarrassingly small: a one-page form that every finding gets written into, with the same fields, every time. This article defines that form — eight fields, one severity rule, one owner column — and how the filled form feeds a fix schedule instead of a wish list. It works as the writing-down layer on top of whatever diagnostic method you already use, whether that is a structured pass with a a small website audit checklist or a looser click-through, and it pairs with the printable what an audit checklist PDF should include if your audits happen on paper.
The principle underneath every field below: a finding is only real once it is written in a form that someone else could act on without asking you a single question.
The Eight Fields Every Finding Needs
One: page URL, with the specific location when it matters — “contact page, form” is actionable, “forms are broken” is not. Two: what you expected, in one sentence — the standard the site is failing, stated plainly. Three: what actually happens, observed rather than interpreted — “the button does nothing on tap” rather than “the button is broken.” Four: who it affects — every visitor, mobile visitors only, visitors using a screen reader — because audience size is the honest driver of priority. Five: how you found it — device, browser, or tool — so the fix can be reproduced before it is claimed fixed. Six: severity, using the three-level rule below, not gut feel. Seven: proposed fix, stated at the level of intent — “add a visible label to the email field” — leaving implementation to whoever owns the site. Eight: owner and date, even when the owner is you; an unowned finding is a finding scheduled for never. Eight fields, one or two lines each, and a finding is done being described and ready to be scheduled.
The Severity Rule: Three Levels, No Fourth
Severity debates sink audit meetings, so the form removes the debate with three levels and definitions. Level one, broken: something does not work — a dead link, a form that fails, a checkout error. Broken is always first, regardless of audience size, because it converts effort into nothing. Level two, blocked: something works but a defined group cannot complete the journey — mobile users who cannot read the menu, visitors who cannot find the phone number; blocked is ranked by audience size, the field four from the list above. Level three, degraded: something works for everyone but worse than it should — slow pages, dated copy, a footer missing the details customers look for, the exact gaps a focused a website footer audit tends to surface. Every finding gets exactly one level; if a finding feels like two, split it into two findings. The discipline pays off when the form fills up: a page of level-one items schedules itself, a page of level-three items is a backlog, and the mix tells you whether the site needs fixes or just polish.
Worked application: a bakery owner audits her six-page site on a Sunday with the form printed beside her. She logs nine findings: one level one — the order form silently drops submissions on mobile — four level two, mostly mobile navigation and one missing allergen page, and four level three, dated photos and slow-loading images. Because the form forced an audience field, she sees at a glance that the level-two cluster is all mobile. That single column reframes the fix schedule: one developer afternoon for the form, a template fix for mobile navigation, and the level-three items slot into her quiet winter weeks instead of competing with the broken form.
From Filled Form To Fix Schedule
The form’s last job is preventing the classic audit ending: findings written, nothing scheduled, form forgotten in a drawer. Two habits close the loop. First, schedule at the top of the form, not the bottom — before writing anything, book the two hours when findings will be triaged, or the audit will join the drawer. Second, triage by level and audience only; resist re-litigating severity in the triage meeting, because that is what the three-level rule was for. Each kept finding leaves triage with an owner and a date or it is deleted deliberately — “not fixing this” is a valid outcome, but only as a decision, never as a default. Run the same form quarterly and the comparisons write themselves: severity mix shifting from ones to threes is what progress actually looks like on a small site, and the form is the only record that proves it. For teams that want a shared vocabulary before the first audit, the Nielsen Norman Group’s ten usability heuristics are the standard starting set — not a replacement for the form, but a checklist of patterns worth hunting for while you fill its eight fields.
If the audit feeds a redesign decision, attach the filled form to the brief verbatim — the severity mix is the strongest one-page argument a stakeholder can read for why the redesign should happen now rather than next year. “Nine findings, four of them level one, all mobile” is a budget conversation; “the site feels dated” is an opinion, and the form is what converts the second into the first without an extra meeting.
The E-Commerce Adaptation: Ten Fields For Shop Audits
On a shop or booking site, extend the form with two fields beyond the core eight. Nine: transaction stage — where in the buy-or-book flow the finding lives, because a broken field on the payment page outranks the identical broken field on a brochure page, and level-one items need this column to be triaged honestly. Ten: revenue proximity — a rough yes-or-no on whether the finding sits within three clicks of money changing hands. The additions keep the form’s discipline intact: same severity rule, same owner-and-date rule, just a sharper sense of which broken things are expensive. And they make the quarterly comparison more honest — a rise in level-one payment-stage findings is a different, more urgent story than the same count spread across brochure pages, and the two columns are what let you tell the difference at a glance.
If you adopt only one field, adopt “who it affects.” Most audit paralysis is really an audience-size argument in disguise, and that single column answers it before the meeting starts.
For teams of more than one, keep the form in a shared document rather than a printed sheet, so triage can sort and filter by severity and owner; the paper version works for a solo audit, but a spreadsheet turns the same eight fields into a backlog that re-sorts itself every quarter.