Regulated products · Guide

UX design for insurance and regulated products

Practical lessons from AXA work: explain complex offers, simplify forms and test mobile comparison without overstating the evidence.

The short answer

Good UX for a regulated product helps people understand what they are choosing, why information is requested and what happens next. Start with the customer decision, involve the people who own product rules, and test comprehension as well as completion. The AXA examples below show delivered components, exploratory work and a concept that remained in the backlog; they support different conclusions.

Design around the decision people need to make

A person considering insurance needs more than a shorter page. They need to understand the situation the product addresses, distinguish options and recognise when they need help. Removing content can make a screen calmer while making the decision harder.

My starting point is to map the questions behind the journey: what is covered, what changes between formulas, what information is needed for a quote, and what a next action actually commits the person to. Product specialists need to verify the meaning of the answers. The design then determines their order, emphasis and presentation.

This is a proposed working method, not a claim that one interface satisfies every requirement in insurance or banking. The cases here concern AXA insurance work; they do not demonstrate a banking implementation or regulatory certification.

Keep the form connected to its purpose

In the AXA decision-tree project, Contentsquare analysis identified low engagement with a form. The existing journey asked people to leave the product page after clicking a CTA. The redesign embedded an easy first question in the page and developed progressive mobile and desktop flows in Figma.

The useful principle is continuity: help someone begin while the reason for the question is still visible. It does not follow that every insurance form should become a conversation, or that fewer visible fields mean less effort. Branches, corrections and handoffs still need design.

The portfolio records improved engagement but publishes no numerical baseline or experiment protocol. I therefore use this case to explain the design decision, not to promise a conversion percentage. The form guide develops the questions to ask before adopting this pattern.

Let research challenge the attractive solution

The AXA comparison matrix explored a swipable mobile alternative to product cards. The intention was to make differences between formulas easier to inspect on a small screen. User testing preferred the existing cards. A second iteration was defined for another validation round.

That result matters: an interactive prototype is evidence that an idea can be demonstrated, while the test challenges whether it solves the user's task. The record does not supply participant counts, task wording or a final V2 result. It supports retaining the challenge and revisiting the proposal, rather than declaring a winning mobile pattern.

For a new project, I would test whether participants can explain a relevant difference, identify a limitation and find their way back after inspecting another option. See the mobile comparison guide for a practical test structure.

Use education and reassurance for different jobs

The awareness examples component used concrete situations to explain why a product might matter. Its ProtoPie prototype was connected to an AXA pet-insurance page. The social-proof work placed a compact trust signal near the top and detailed reviews further down the page.

These components answer different questions. An example helps explain a need; a review describes someone else's experience. Neither should substitute for the product's scope and limitations. The social-proof record reports better engagement and a slight CTA click-through lift without the underlying figures or causal controls.

The quiz concept offered another route into awareness through questions and feedback. It reached sizing and backlog, then was parked for capacity. It illustrates a design direction, not a launched learning experience. The product-page guide connects these roles.

Make the review concrete before release

Bring an actual journey to the review: a first-time visitor, an uncertain answer, a correction and an unsuccessful submission. Agree which product statements require specialist approval and record who owns changes when the offer evolves.

For accessibility, W3C's form guidance explains how labels identify controls and remain associated with them. That informs implementation; a prototype annotation alone cannot establish conformance. Test the implemented flow with keyboard and assistive technology as well as touch.

Treat analytics and research as complementary. A rise in starts can coexist with confusion later in the journey. Define the intended customer understanding, instrument meaningful steps, and document the device, language, date and sample behind any conclusion.

Your brief

Tick what you can already answer. Unknowns are fine: just name them.

Questions & answers

Does simpler UX mean removing insurance information?

It means making relevant information understandable and available at the decision that needs it. Any removal or reinterpretation of product information needs review by the people responsible for that content.

Were all these AXA concepts launched?

No. The records describe implemented awareness and review work, form prototypes with an engagement outcome, a comparison exploration redirected by user testing, and a quiz parked at backlog stage. Those statuses should remain distinct.

Can these patterns be applied to banking?

The questions about comprehension, correction and evidence are useful starting points. Banking product rules and user tasks require their own discovery and validation; these insurance cases do not establish that a banking journey will work.

Sources & context

  • AXA: Decision Tree Form Redesign (2023)

    Documents the separate-form handoff, Contentsquare analysis and embedded decision-tree prototypes. The record describes improved engagement without a published numerical baseline or attribution method. First-person project account; the underlying research and analytics are not published here. Checked .

  • AXA: Mobile Comparison Matrix (2024)

    Documents a Figma exploration, preference for the existing cards in user testing and a second iteration awaiting further validation. It does not establish successful rollout. First-person project account; the underlying research and analytics are not published here. Checked .

  • AXA: Awareness Examples Component (2024)

    Documents an interactive ProtoPie prototype and connection to a live pet-insurance journey at the time of the project. Current AXA production behavior was not retested. First-person project account; the underlying research and analytics are not published here. Checked .

  • AXA: Social Proof Reviews (2022)

    Documents a review plugin, two reassurance placements and qualitative engagement/CTA observations. No experiment duration, sample size or numerical uplift is published. First-person project account; the underlying research and analytics are not published here. Checked .

  • AXA: Product Quiz Component (2025)

    Documents a designed and sized awareness quiz parked for capacity reasons. The concept is not evidence of production use or measured customer outcomes. First-person project account; the underlying research and analytics are not published here. Checked .

  • W3C WAI: Labeling Controls

    Primary implementation guidance fetched on 10 November 2026; page states updated 13 May 2024. Used for the association between labels and controls, not to certify any portfolio implementation. Checked .

Make a complex insurance journey easier to use

Share the product, the customer decision and the step your team needs to improve. We can define a focused review and a testable next iteration.