Let's Connect

Let's Connect

The Zebra

Reducing funnel drop-off at scale

Product Design • Conversion Funnel • A/B Testing

applicant conversion

+12%

funnel drop-off

-15%

revenue per user

↑RPU

THE CONTEXT

THE ZEBRA

The Zebra compares over 200 insurance companies to surface personalized quotes. The funnel was the critical path — but it was a black box. Users didn't know where they were, how far they'd come, or how much was left.

THE PROBLEM

Users entering The Zebra's insurance funnel had no visibility into how long the process would take. Without progress signals, they felt overwhelmed and dropped off before completing the questions — meaning no quotes, no revenue, no match.

SCOPE & CONSTRAINTS

WHAT WAS IN SCOPE

Device: Mobile only
Insurance line: Auto funnel
Page range: Top of funnel (after ZIP entry, Q1) through end of funnel (last question before results)
Excluded: Pre-populated SEO/SEM pages and returning users

WHAT SUCCESS LOOKED LIKE

Primary: Increase in User-to-Applicant conversion — users who enter the funnel and reach the results page
Secondary: RPU (Revenue Per User)
User outcome: User knows their place in the funnel, how much they've answered, and what's coming next

DESIGN HYPOTHESIS

IF we set user expectations for how long the funnel is BY adding a progress indicator throughout, THEN we will increase user-to-applicant conversion BECAUSE the user knows what to expect.

DESIGN ITERATIONS

1 / 4
1 / 4

PROPOSED DESIGN SOLUTIONS

A/B TESTING

The variant was shown to 50% of auto insurance traffic over several weeks — or until the funnel reached a statistically significant sample size. Testing was run through Eppo, with the primary lens on drop-off rate via User-to-Applicant Conversion.

CONTROL

No progress indicator. Users enter the funnel with no sense of length, completion, or what's next.

VARIANT

Progress indicator added throughout the funnel. Users can see where they are and how much remains at every step.

FINAL SCREENS

IMPACT & OUTCOMES

+12%

applicant conversion

↑RPU

revenue per user

−15%

funnel drop-off

5,000+

users in test sample

NEXT STEPS

Analyze A/B results for statistical significance and determine whether lift is strong enough to expand the pattern.

Iterate on the progress indicator design based on what the test data reveals about user behavior at different funnel stages.

Expand to other insurance funnels — home and bundle — if the auto funnel results prove out the hypothesis.

WHAT THIS PROVES

I understand how to move the needle in a real product

I understand how to move the needle in a real product

This wasn't a redesign — it was a focused, testable intervention in a high-traffic system. I identified a clear user problem, formed a measurable hypothesis, designed within tight constraints, and validated the solution with data. That's how I work: impact-first, constraint-aware, and always ready to prove it.

Let's Connect

SCOPE & CONSTRAINTS

WHAT WAS IN SCOPE

Device: Mobile only
Insurance line: Auto funnel
Page range: Top of funnel (after ZIP entry, Q1) through end of funnel (last question before results)
Excluded: Pre-populated SEO/SEM pages and returning users

WHAT SUCCESS LOOKED LIKE

Primary: Increase in User-to-Applicant conversion
Secondary: RPU (Revenue Per User)
User outcome: User knows their place in the funnel, how much they've answered, and what's next

DESIGN HYPOTHESIS

IF we set user expectations for how long the funnel is BY adding a progress indicator throughout, THEN we will increase user-to-applicant conversion BECAUSE the user knows what to expect.

1 / 4

Interested in working together?

Let's build something meaningful.

Let's Connect

Reducing funnel drop-off
at scale

Conversion Funnel • A/B Testing

The Zebra

applicant

conversion

+12%

funnel

drop-off

-15%

revenue

per user

↑RPU

THE CONTEXT

THE ZEBRA

The Zebra compares over 200 insurance companies to surface personalized quotes. The funnel was the critical path — but it was a black box. Users didn't know where they were, how far they'd come, or how much was left.

THE PROBLEM

Users entering The Zebra's insurance funnel had no visibility into how long the process would take. Without progress signals, they felt overwhelmed and dropped off before completing the questions — meaning no quotes, no revenue, no match.