Skip to content
All work Say hi

Case study · Waah Jobs (OLX Group)

Onboarding, not screening: 61.7% to 82.4% in three weeks

Jobseekers were leaving Waah Jobs before they ever saw a job. I rebuilt onboarding around one question: what’s the least we need to ask before showing someone a relevant job? Conversion rose from 61.7% to 82.4%.

Role
Senior UX Designer
Timeline
3 weeks
Team
Me, with Shubhangi Jadhav (Design Manager) and Shital Khairnar (Senior PM)
Skills
Interaction design · User flows · Prototyping · Product thinking
Waah Jobs onboarding conversion from 61.7% to 82.4% in three weeks, by asking only what a job list needs: onboarding, not screening. Beside it, the splash screen, the new phone-number landing screen and one-tap Truecaller verification.
  • 61.7 → 82.4%Onboarding conversion
  • −36%Worst per-screen drop, eliminated
  • 3 weeksFrom problem to shipped flow
01

The drop-off

Every screen lost more people

Our app research had named onboarding as the biggest leak. The flow asked for 20+ details, copied from white-collar job boards whose users have CVs ready, before showing a single job. For blue-collar jobseekers it was a wall, and the funnel showed it.

By the work-details screen, 36% had left without seeing a job.
The old flow: long forms, email and passwords, before any job.
1 / 2
02

What we knew

We already knew more than we thought

Before designing anything, I went back through past A/B tests and interviews for this exact flow. Four things stood out:

  • Phone beats social loginUsers feared sharing social accounts and preferred a phone number with OTP.
  • Trust is fragileJob fraud is common, so asking a lot up front read as a red flag.
  • Small steps beat wallsShort contextual pages tested better than one long form.
  • Email is a dead endOnly 6.4% had an email they actually used, and everyone forgot passwords.
03

The reframe

Onboarding isn’t screening

In a brainstorm with my PM and design manager, I questioned the requirement instead of the screens. We needed all that data to match a perfect job, but did we need it before people even understood the product?

That became the rule for the whole flow: ask only what’s needed to show a relevant job list, tune search to work from those few answers, and collect the rest later, when its value is obvious.

The new flow: only what a relevant job list needs.
04

Three moves

Three moves did the work

I worked through flows and low-fidelity screens before any UI, then tested a clickable prototype before build.

Lead with trust: verified jobs, no hidden charges, one clear action.
Remove the keyboard: Truecaller one-tap login, profile pre-filled.
Feel shorter: tap-based steps instead of typing, ending on a success screen.
1 / 3
  • Lead with trustA Made-in-India splash and a landing screen that answers the fraud fear first.
  • Remove the keyboardTruecaller one-tap login skipped typing and pre-filled the profile.
  • Make it feel shortAnything a job list didn’t need moved out; the rest became quick taps.
05

Results

61.7% to 82.4%, and the rest of the profile later

After launch, onboarding conversion rose from 61.7% to 82.4%. For the profile details we no longer asked up front, we mapped nudge points later in the journey, at moments where people can see that more information means better jobs.

Profile nudges: asked at the moment the value is obvious.
06

Lessons

What I’d take into the next role

Three lessons I’m taking with me.

  • Question the requirement, not the screenThe biggest win came from asking why we needed the data, not from a better form.
  • Borrowed patterns need a fit checkA flow built for white-collar CVs failed people who don’t have one.
  • Time to value is the metricGetting people to a relevant job fast mattered more than a complete profile.