Babylon Health — Improving the First Encounter Experience

This work is password‑protected

Don’t have the password? Email me.

Incorrect password — try again.

TL;DR

Babylon’s US registration felt like a helpline transfer queue. We rebuilt it around insurance, consent and families, and every metric moved.

93%

from 60%

Sign-up completion

86%

from 22%

Insurance check completion

3×

since launch

Appointment bookings

The problem~35% abandoned at sign-up alone — and insurance eligibility was checked last, after all the effort was spent.

The researchThree pain points: no trust, an opaque process, families left behind — in a journey built for individuals, when most eligible members had dependents.

The goalHelp members create accounts and pass the insurance eligibility check, so they and their families reach the right services — shipped in phases to learn from every release.

The solutionExplain the service up front, check eligibility first, onboard the whole family, end on a welcome appointment.

01 · The challenge

Registration that felt like a helpline transfer queue

Ever been bounced between five departments on a customer helpline? Babylon's US onboarding felt the same. Each step in the journey belonged to a different internal team — marketing, legal, security, clinical — and users could feel the organisational seams.

Problem

Registration was inefficient for onboarding and verifying eligible members, with no intuitive path for managing dependents’ health.

The 2020 US registration journey, mapped step by step against the funnel data.

Outreach → sign-up

77% of prospective members never responded to the sign-up outreach

First step

~35% dropped off at the sign-up step alone

Families

10% sign-up rate for children without an eligible adult in the household — 40% lower

Those who did start found a journey that leaked at every hand-off:

The old registration journey mapped step by step, showing team ownership and drop-off at each stage — The 2020 journey mapped — every step owned by a different team, with the insurance check buried in appointment booking

02 · The research

Understanding why people never finished signing up

User decision map across six stages — triggers, awareness, validate, try, purchase and keep using — with the touchpoints and product features met at each stage — User decision map

Opportunity solution tree structured by the Fogg model: motivation, ability, trigger — Opportunity solution tree

Each pain point became a design question:

Pain point 01

Lack of trust & understanding

Users were unsure how the service worked with their insurance, questioned the value on offer, and wanted clarity on legal compliance before committing.

HMW increase users' motivation and knowledge of the services so they onboard with confidence?

Pain point 02

A long, opaque process

The journey asked too much of users while explaining too little — a tedious form with no sense of progress or purpose, leading to confusion and abandonment.

HMW offer a secured, faster path to the right healthcare services?

Pain point 03

Families left behind

Onboarding centred on the individual adult, neglecting parents and carers who needed to manage the health of children and dependents.

HMW help adults onboard eligible dependents to the service with ease?

03 · The design

Bridging gaps to improve the user experience

Slide 02 — Formulate a coherent theory from research inputs to provide an implicit intro guide. The intro carousel on a phone standing on a rock.

Slide 03 — Communicating how to onboard dependents upfront; dependents onboarding rate increased about 15%. The dependent-care-covered screen on a phone in hand.

Slide 04 — Reduce confusion by introducing an adult-or-guardian badge when collecting personal data. The what's-your-name step on a phone resting on a rock.

Slide 05 — Provide better context and give autonomy back to users. The add-eligible-dependents screen on a phone held in two hands.

Slide 06 — Simplify adding dependents while ensuring appropriate security: names partially masked before verification, each child skippable.

Slide 07 — Introduce a welcome appointment as a clear next step at the end of onboarding, on a phone in hand.

Ten screens in two rows: the registration steps — email, verification code, name, date of birth and the matched health plan — shown first in the light theme and again in the dark theme. — Every step designed for both light and dark themes

04 · The outcome

Every metric moved

Sign-up completion

60%→93%

Modularised signup · improved consents · value props · leveraging insurance data

Insurance check completion

22%→86%

Check moved into onboarding · family prompts · better error states · third-party data

Children onboarding

13%→75%

Added within insurance check flow · "how to" guidance · express children onboarding

Appointment bookings

since launch3×

Next-best-action guidance · welcome appointment · incentives at the right moment

05 · How I decided

Three calls that mattered

Friction over fuel

The A/B test said the quiet part: faster onboarding lost to trust-building. We kept the intro guide and spent the friction where it earns belief — removing emotional friction beat removing steps.

Kill the magic link

Auto-retrieving insurance data was technically easy and emotionally wrong — participants found the unexpected disclosure invasive. Autonomy became the rule: explain, play back, and ask first.

Design around the legal minimum

US healthcare law requires three pieces of personal information at minimum, so registration collects only those three. Everything else waits until it's needed.