[ case study ]
From onboarding users,to activating a weekly
networking habit
Redesigning Veezo's first-time experience around its AI agent - so the first conversation becomes the first step in a recurring networking behavior.
Context
Veezo helps people make their professional network more useful. Its AI agent adds a conversational layer on top of contacts, connections, calendar and memories - helping users find relevant people, organize relationships, and act on networking opportunities.
As the agent became a central part of the product, onboarding had a new job.
It no longer needed to explain Veezo. It needed to get users to experience its value.
The problem with the first onboarding
The original onboarding followed a familiar mobile pattern: introduction screens, personalization questions, and then access to the app.
It helped users understand what Veezo was - but left them to figure out what to do with it.
Once inside, users still had to:
- explore the interface.
- discover the AI agent.
- understand its value.
- and decide how networking with Veezo should fit into their routine.
The onboarding ended exactly where the important behavior needed to begin.
The goal: activation, not completion
The question was no longer:
How do we get users through onboarding?
It became:
How do we get users to experience a behavior worth repeating?
We defined the real activation moment as behavioral:
The user understands that Veezo can regularly introduce them to relevant people - and commits to making space for that interaction every week.
The old onboarding ended with access. The new onboarding needed to end with momentum.
The shift: onboarding through use
Instead of introducing the AI agent after setup, we made it the first interaction after sign-in. The agent became the onboarding.
The agent became the onboarding - guiding users from inside the product instead of explaining it from the outside.
Rather than describing what Veezo could do, it began doing it with the user: learning their context, clarifying their networking goal, and moving them toward a meaningful first action.
We were inspired by conversational-AI products that reduce traditional onboarding and bring users directly into the core interaction.
But we did not want to simply remove onboarding. Veezo still needed to guide users toward a specific behavior.
Different starting points, one destination
The original onboarding asked generic personalization questions upfront, before users had much context for answering them.
The AI agent allowed personalization to happen progressively - asking only what was needed, when it was needed.
Users could arrive in very different states. Some came through a Circle invite, some without any group context. Some connected LinkedIn immediately, while others delayed setup. Some knew exactly who they wanted to meet; others needed help defining their goal.
We mapped these entry states so the conversation could adapt while still guiding everyone toward the same outcome:
Meet relevant new people on a recurring basis.
Networking is valuable, but vague and easy to postpone. The onboarding therefore turns it into a concrete behavior: define a goal, choose a regular time, and let Veezo introduce someone relevant each week.
Different paths. One destination. One behavior worth repeating.
The activation
Build context
Connect relevant sources (Google, LinkedIn, phone contacts) and understand who the user already knows.
Clarify intent
Understand who they want to meet, why, and what they can offer in return.
Create the first match
Use that context to surface a relevant person the user could meet.
Turn it into a recurring behavior
After the first match is approved, establish a regular networking slot and introduce the broader Veezo ecosystem.
Once activated, Veezo works in the background toward the user's networking goal and brings them a relevant person to meet each week. Around that core behavior, users can organize their network, save context about relationships, schedule meetings, and explore warm introduction opportunities through Circles.
The weekly introduction is the anchor. The rest of the product supports it.
Design outcome
The redesign shifted onboarding from a setup sequence into the first real product experience.
Instead of asking every user to complete the same introductory flow, the AI agent adapts to their context, helps clarify their intent, and leads them toward a meaningful first connection.
Before: onboarding ended when the user entered Veezo.
After: onboarding ends when the user understands why they should come back.
The goal wasn't to teach people how to use Veezo. It was to give them a reason to use it again.
What happened next
The redesigned onboarding was completed and ready for launch, but the company shut down before the new flow was released.
Because of that, I wasn't able to validate its impact with live users.
Had it launched, I would have measured:
- first meaningful match completion
- meeting approval or scheduling
- conversion to a recurring weekly networking slot
- week-one return rate
- drop-off points in the conversational onboarding
The design was complete, but the behavioral hypothesis remained unvalidated.