The UX practice for insurance & embedded protection

Insurance UX fails in specific ways: terms nobody read before paying, uploaded photos of a broken phone that get dropped from the system, and an adjuster who can’t trust what the screen tells them. I find those failures so you can fix them before they cost your business money and loyalty.

12 years in production UX · In-house at Assurant · Enterprise design for a Fortune 50 telecom · English / Spanish · US Central hours

Book a Call See the work ↓

Why insurance

Insurance holds an abundance of UX wins, but it needs expertise to uncover and implement them.

In my time working in insurance, I’ve noticed that it tends to be an oft-neglected industry in terms of usability. Yes, there are some newer services that purport to offer a more design-forward experience, like Lemonade, but I find that the experiences with polish tend to be a façade over the more gritty details—the edge cases, the points where a customer needs to speak to a rep, or the ability to actually have a good understanding of coverage before purchasing the policy in the first place.

On top of that, I’ve had the opportunity to spend a few years working on backend claims systems, and there have always been tremendous opportunities for improvement in that realm, which has impacts for both customers and employees—from meeting SLAs and disclosure requirements, to keeping things consistent. Of course, improving these things affects the bottom line for these businesses as well.

Many designers might look at the insurance industry as unsexy compared to, for example, the tech companies operating out of Silicon Valley. In my view, however, that group is discounting a major opportunity for impact, and to improve the lives of a massive industry, and an even more massive group of people.

The figure below shows a demonstration of bringing several different claims systems together under one unified application. Less uncertainty, more reliability, easier handoff between teams, improved SLA times, and a happier customer.

FIG. 01 SHARED TIMELINE QUEUE EXCEPTION SUMMARY ▮▮▮▯
FIG. 01 — Claims systems, consolidated. On the left, the independent APIs from previous systems feed into the unified view for the user. Dates and actions taken from the different systems are combined into the shared timeline at the top. Different symbols are used in addition to different colors for accessibility. An AI summary is implemented at the bottom so that the user can quickly determine what their most urgent action items are at a glance.

Services

Three engagements, each at a different level for different needs.

Each level brings the same expertise: I build designs that hold up for your customers on their worst day, and for your business on every other one.

In this domain, the work that serves the customer and the work that serves the numbers turn out to be the same work: a better user journey means fewer calls, improved accessibility, happier customers, higher ratings, and less of your budget spent on fixing the aftermath of a broken system.

Pick by situation: a journey you suspect is broken takes the audit; a design team that keeps stalling on insurance questions takes the retainer; a work stream stuck behind an unfilled headcount takes the embedded engagement. For whatever you need, you’ll receive the expertise it takes to know this domain on sight.

The UX Audit

$5,000 flat · 2 weeks

A UX audit offers a double-win for your business: an improved user experience, and saving your bottom line in doing so. I will be meticulous in examining every aspect of your most important user flow to ensure optimized usability, and it will be reflected in the final report.

The report will also include expected improvements in the metrics your analytics provide, and how to measure it. If you have no analytics implemented yet, I will explain how to implement them as a part of the report.

You get:

  • An end-to-end deck: highlights and top wins first, then every finding in full
  • A prioritized fix list, split into ship-this-quarter quick wins and structural work
  • Trust and compliance flags: where language, disclosure, or state variation creates risk
  • Three sessions — a 30-minute kickoff to align on expectations, a 90-minute meeting to walk the recommendations, and a 90-minute implementation session with your dev team

The deliverable

An end-to-end deck explaining the results after auditing your most important user flow. Highlights and top wins shown first, with a comprehensive breakdown of each finding and recommendation within, including the issue, category, explanation of the harm it’s doing, the suggested remedy, and analytics suggestions to measure the impact.

Includes screenshots, example code snippets illustrating a fix, and any applicable accessibility or UX heuristic violations. The deck can be anywhere from 25 to 80 slides long depending on the number of steps and issues in the flow.

What I need from you

  • Access to your full site flow, with login details if required
  • Design system & guidelines
  • A quick 30 minute discussion at the beginning of the audit to align on expectations.
  • A 90 minute meeting at the end of the audit to discuss recommendations.

What happens after the audit

Throughout my career, I’ve worked with development teams to bridge the gap between design and implementation. Additionally, as someone who has hand coded apps in the past, I understand what is and is not technologically feasible. Once the audit is finished, I will attend one 90 minute implementation session with your dev team to make sure that the recommendations from the audit don’t get lost in translation, with an eye toward how the recommendations will live in your codebase.

No discovery phase, no scope negotiation, no surprise invoice. The kickoff is expectation alignment, not discovery. One journey, two weeks, a punch list your team can ship.

Set up a quick call here to get started.

One audit slot at a time. Less than a week of a senior designer’s loaded cost, and a fraction of an agency discovery phase. Read a published teardown →

Advisory retainer from $3,500/mo

For teams who already have capable designers, but who either stall on insurance questions, or are already at capacity on other projects. If your team needs answers, but doesn’t necessarily need more hands-on execution, then there’s no reason to buy hours that you won’t use. The retainer’s advantage is that it helps prevent expensive mistakes and unblocks decisions, without paying an extra person to sit in your daily standup.

You get:

  • Async access to me for the questions that stall your team, answered within one business day.
  • A design-review pass before anything ships: flows or screens flagged with any issues found, and with the reasoning written out.
  • Answers with receipts: every recommendation points at the expectation or document behind it, so your designers learn the why—not just my verdict.
  • A weekly 1-hour call to work through whatever stalled hardest that week, with whoever on your team is closest to the problem.
  • Month-to-month coverage, with thirty days’ notice either direction. If you’re predicting a quiet month, pause with one call instead of paying for that gap.

Embedded engagement from $8,000/mo

I can lend a hand if your team needs more help with work that has an end date, or with work that increases mid-project. Maybe your team lacks design leadership to manage a new design hire. Or, you just need help implementing the audit. For any of your projects that need an expert who can take responsibility for the UX, end-to-end, I’ll be part of your team, there to assist any junior designers, and to debate with the engineers about executing the design.

You get:

  • Kickoff within a week of signature.
  • One workstream owned end to end: problem definition, user flows, specs, and implementation support alongside your engineers until it ships.
  • Roughly two days a week inside your team.
  • A written handover at ramp-down, including the decision log, the reasoning behind it, and the open questions.
  • Three-month minimum, rolling after, with the ramp-down month planned deliberately, not improvised in the final week.

Contracting

This is a consultancy with a team of one, so you will be working with me directly. In the case of an audit, the contract will be for the duration of two weeks. Payment is 50% upfront / 50% on completion. For larger fixed-scope projects, 30–50% upfront, then milestone payments. Retainers invoiced monthly. Pausing an advisory retainer requires 30 days’ notice. Resuming will depend on availability.

I take a maximum of three retained clients.

Selected work

Production work, abstracted on purpose

Visuals are abstracted recreations with fictional data — client work stays confidential. The thinking is the exhibit.

One system

Startup design system — RealTheory

Engaged as an outside consultant for end-to-end product design. Built the company’s first design system, led the Sketch-to-Figma migration, and owned analytics from tool selection to feature-performance metrics.

Claim status Received → In review

White-label claims platform — Assurant

Claims experiences that telecom and device partners deploy under their own brand. Also at Assurant: a complex Microsoft Dynamics CRM redesign, applying heuristic analysis and usability testing to streamline internal claims workflows.

Schedule Applications Curriculum

Career-transformation platform — The Home Depot

One internal platform for applications, scheduling, curriculum, and role-based access — designed around personas and journeys for students, instructors, and administrators.

FIX LIST SHIP THIS QUARTER

Published teardown — Extend attach flow

What the audit actually produces, run in public on two live storefronts: twenty findings against real screenshots and DOM measurements, and a fix list split by whether the platform or the merchant owns the row.

AI in production

Shipping AI into your product? Design the trust layer first.

Insurers are putting AI into quoting, claims triage, and service — and discovering the hard part isn’t the model, it’s the interface: how confidence is displayed, what happens when the system is wrong, when and how a human takes over, and how an automated decision gets explained to a policyholder or a regulator.

That’s interaction design work, and it’s where I focus my ongoing research: oversight patterns, error and fallback states, and explainability in regulated flows.

Two of the principles that fall out of that work: the summary is a trust surface, not a text box — confidence displayed, sources linked, the raw record as fallback. And sources stay drawn as redacted — the confidentiality constraint is the aesthetic, here and everywhere else on this site.

Confident ≥ 70% confidence

87% confidence

Customer reports accidental drop resulting in cracked screen and rear damage. External report confirms damage exceeds policy limit. Estimate generated for customer review.

Sources (5)
  • ↳ Claim record
  • ↳ Adjuster note
  • ↳ Customer statement
  • ↳ External report
  • ↳ Photos (3)
Take over this summary
Uncertain 30–69% confidence

42% confidence

Limited information about this claim. The summary may be incomplete or inaccurate.

Confidence is low because
  • Photos missing
  • External report not received
  • Statements are inconsistent
Sources (2)
  • ↳ Claim record
  • ↳ Adjuster note
Take over this summary
Unavailable Model unavailable

No summary generated

The source record is below, unchanged. Nothing is hidden while the model is down.

Write the summary
MODEL GENERATES SUMMARY HUMAN REVIEWS OR EDITS DECISION RECORDED
Every state ends the same way — a person decides, and the decision joins the record. Fictional data throughout.

About Sam

Portrait of Sam Eastham

I’ve been passionate about UX even before I knew the term for it—since I was young, I’ve had the idea in my head that many times, when “mistakes” are made by the user of the system, it’s the system that has failed—not the user. Once I learned about the field itself, I have loved bringing design wins to teams: increasing efficiency, saving money, reducing mistakes, and making the lives of the users better, all thanks to user centered design.

User experience is always on my mind—most recently, as an example, I was fascinated by the ergonomic design of the Ninja Creami while making my own ice cream. The Design of Everyday Things was my first UX book, and Don Norman is one of my heroes.

When no one’s paying me, I design anyway — lately Future Retro, a full design system built for its own sake.

Book a Call

Contact

Not ready for a call, but still have questions?

I’m open for UX discussions any time via email. Send me your information and what you think might be underperforming in your product.