Jeet Bania
Back to blog

Design

Designing InCore's onboarding, three times over

The first version was correct and nobody used it. The second was fast and nobody trusted it. Here's what the third one got right.

Author
Jeet Bania
Published
June 18, 2026
Read time
6 min read

The problem

Every engineering team we onboarded to InCore asked the same question in week one: "where do I actually start?" The product had forty capabilities and one entry point — a dashboard that treated a first-time licensing lead and a five-year power user identically.

Our first attempt fixed this with more structure: a guided checklist, progress bars, the works. It was correct. It was also completely ignored — completion rates sat at 11% after three weeks.

What changed

The checklist assumed people wanted to be taught the product. What they actually wanted was to finish the one task that brought them there — usually licensing a specific core — and see everything else later, on their own terms.

Onboarding isn't a tour of the house. It's directions to the one room they're trying to get to.

  • Collapsed forty capabilities into three entry intents, chosen up front
  • Replaced the checklist with a single contextual next-step, one at a time
  • Let the dashboard stay empty until there was something real to show

Takeaways

11% → 64%
Onboarding completion, week one
3rd
Attempt that actually shipped
9 days
Median time to first license

The lesson wasn't "simplify." It was that a guided tour and a shortcut solve two different problems, and I'd spent two rounds solving the wrong one.

Next post
What nobody tells you about handoff

Still here?

Work with me.

Jeet Run

Dodge client chaos

SCORE

00000

BEST

00000

↑ Space / tap to start · dodge client chaos
© 2026 Jeet Bania · Definitely not my first draft.Built with love, caffeine, and too many Figma frames.
Switch to dark mode