top of page

Enhanced Affordability

William Hill & 888 Occupation selection

Redesigning how William Hill &888 collects occupation data from customers, turning a friction-heavy compliance gate into a clear, trustworthy two-step flow.

My Role

UX Designer/Researcher

Status

Live: monitoring in progress

help76.png
THE PROBLEM

A compliance requirement with a UX problem

Under UK gambling regulations (Enhanced Affordability checks), William Hill is required to collect and verify customers' occupation and employment status. The existing solution, a single-field pop-up with threat-led copy, created confusion, drop-off, and distrust.

The old design warned customers "we may restrict your account" with no context, a dismissible "Do it next time" escape, and no employment status step, just a raw job title field. It prioritised compliance over experience, and risked alienating customers at a sensitive moment.

BEFORE

  • Single job title field, no employment context

  • Threat-led copy, fear-based framing

  • Yellow CTA, off-brand feel

  • Dismissible — "Do it next time" escape

  • No error states or validation feedback

  • No competitor precedent applied

AFTER

  • Two-step flow: employment status → job title

  • Reassuring, account-benefit framing

  • On-brand WH dark UI, consistent CTA

  • Required completion, no avoidance path

  • Full error state coverage, 3 failure modes

  • Patterns informed by 7 competitor audits

RESEARCH APPROACH

Before designing, I audited how other regulated and high-trust apps collect occupation data, looking at flow structure, copy tone, field patterns, and error handling.

Screenshot 2026-05-08 at 09.35.42.png

Competitor benchmarking 7 products

These weren't chosen because they're gambling competitors, they were chosen because they face the same core UX challenge: collecting sensitive personal data from users who didn't ask to provide it, in a regulated context where completion is non-negotiable. Benchmarking within gambling would have anchored the redesign to a low baseline. Going outside the sector to fintech, where data collection UX is genuinely more mature,  meant the new design could leapfrog industry norms rather than match them.

Revolut

Fintech

Employment status first, then free-text with tag chips. Progressive and low friction.

Monzo

Fintech

Multi-step: status → industry → job title. Detailed but well-paced with clear progress.

Chase

Fintech

Inline contextual prompt.
Framed around account benefits,
not compliance threat.

Starling

Fintech

Income source selection with illustrative icons. Friendly, non-intimidating framing

Opal

Regulated

Type-ahead search with immediate results. Clean and fast, no radio list needed.

Kraken

Regulated

Occupation dropdown with search. Mandatory KYC, completion non-negotiable

  • All leading fintech apps separate employment status from job title — never a single field

  • Trust-building copy ("helps us serve you better") consistently outperformed restriction-led warnings

  • Type-ahead search with a pre-populated radio list reduced input effort and errors across all audited apps

  • No competitor offered a "do it later" escape on a compliance-required field — completion was mandatory

A two-step flow built for trust and completion

The redesigned flow breaks the task into two focused screens, reducing cognitive load and making each step feel purposeful rather than bureaucratic.

 

  • Entry screen reframes the task as account maintenance, removing anxiety before the form begins

  • Employment status step gives context that makes the job title step feel logical, not intrusive

  • Job title search uses type-ahead with radio selectio, matching Revolut and Opal patterns users already trust

  • Back navigation returns to employment status, not the app , keeping users within the flow

ChatGPT Image May 8, 2026 at 09_14_08 PM.png
Entry

"Keep your account in great shape"

Step 1

Employment status

5 radio options

Step 2

Job title search

type-ahead + radio list

Done

"Thank you for submitting"

​

EDGE CASE

Full error state coverage, three failure modes

A key part of my contribution was mapping every failure path and designing distinct, actionable error states for each scenario.

Validation error

Invalid occupation, inline toast prompts correction without losing entered data and

prompt for "Other' if not found.

Generic failure

"Something went wrong." Full-screen error with Try again + Back recovery path

Save failure

"Your occupation wasn't saved." Distinct from submission error, same recovery, different cause

ITERATIONS

The flow was refined four times in Figma, each round sharpening copy tone, CTA hierarchy, and the employment status to job title step relationship.

​

Designs were built for both the William Hill light & dark UI and the 888 dark , and Mr Green lighter brand context.

ChatGPT Image May 8, 2026 at 09_01_55 PM.png

© 2026 by Shiera Bester.

bottom of page