Our verification and compliance flow hadn't been benchmarked against the wider form-design landscape in some time. I ran a structured competitive UX analysis to identify where the experience lags, where it already leads, and which patterns have become baseline expectations users now bring with them from other products.
Method & Scope
Our closest direct competitors sit behind paid logins with no public demo, so a like-for-like audit wasn't possible. Instead, I built a proxy comparison set, chosen because each reference is strong exactly where our flow tends to be weak: two general-purpose form builders (Typeform, Jotform) representing the polished, conventional end of form design, and two public-sector benchmarks (NHS 111, HMRC Self Assessment) built on the GOV.UK Design System, representing long, high-stakes forms designed for a mixed-confidence audience.
I audited each product, including our own, against eight shared categories: UI & visual design, elements & field options, layout & navigation, content & language, input/validation & errors, upload & evidence handling, trust & reassurance, and accessibility — scoring each 1–5 to make patterns easy to spot at a glance.
| Product | Avg score |
|---|---|
| Our platform | 1.9 |
| Typeform | 2.9 |
| Jotform | 3.0 |
| NHS 111 | 3.1 |
| HMRC Self Assessment | 4.5 |
Key Findings
Conventions users now expect
Three patterns appeared across three or more products — meaning they're no longer differentiators, they're baseline expectations: inline validation as the user types rather than only on submission, unmistakable visual hierarchy around the active question, and consistent contrast and focus states throughout.
Where we already lead
The benchmark surfaced a genuine strength: a dedicated, purpose-built overlay for per-question guidance. None of the four comparators have an equivalent mechanism — Typeform and Jotform offer no per-question help at all, and NHS 111 and HMRC rely on inline text instead. The gap wasn't the mechanism itself, it was how well it was being used — guidance often restated the question rather than adding value.
Where we're behind
Three themes, in priority order:
Recommendations