Redesigning an AI-powered landscaping mobile tool for seamless customization, first-time trust, and feature discovery.
Neighborbrite is an AI-powered landscaping platform that helps users transform their outdoor spaces with personalized design support. Its mobile-first approach lets people visualize and plan landscaping projects straight from their phone — home improvement made accessible, engaging, and efficient.
Ran user interviews and usability tests to surface friction in the AI customization flow, then translated findings into a redesigned mobile experience.
The redesign cut early drop-offs, surfaced premium AI features, and onboarded hundreds of thousands of new mobile users.
Three patterns bled first-time users before they ever reached the AI features.
Login wall first
5/8 hit a login wall before they understood what the tool did.
Customize hidden
8/8 overlooked Customize — Reimagine took all the visual weight.
6/8 first-time users didn't know how to start — no walkthrough, no demo.
Three targeted moves — delay the login, balance the AI buttons, and onboard with a demo.
Redesigned flow
Login moves deeper into the flow — users explore before committing.
Customize and Reimagine share the spotlight — no more buried AI value.
A "Watch Demo" entry teaches the AI flow before the user has to try.
A one-on-one with the founder to identify where the AI customization flow was leaking users.
Three business goals from the founder, translated into the questions that drove every session.
Map AI friction — identify friction in the AI inspiration and customization journey.
Where do users hit friction when generating landscape inspiration with the AI tools?
Free vs premium — uncover blockers to free vs. premium feature engagement.
What customization features do users expect, and what shifts them from free to paid?
Trust & onboarding — evaluate trust signals and onboarding gaps for first-time users.
How do first-time users build trust before committing to signup?
A qualitative research plan to dig into the friction the stakeholder briefing surfaced.
Eight participants ran a 45-minute session on Zoom — scenario-based exploration and a post-test survey closing every one.
8 participants · 4 professional landscapers, 4 homeowners · task-based exploration with a post-test survey closing every session on Zoom.
Scenario"Imagine you just moved into a new home with a large yard and you're looking for inspiration." Five tasks walked each participant through the full flow.
Task breakdownEight sessions surfaced three things the platform already does well — and three places it leaks users.
Strengths to protect in the redesign — visual identity, AI speed, and the plant database all landed.
7/8 praised the modern, clean interface as one they'd trust with their home.
6/8 loved how quickly the AI generated landscape ideas.
5/8 valued the quick access to species info, care tips, and growth details.
Three patterns repeated across every session — each disrupted the mobile experience.
5/8 hit a login prompt before they understood what the tool offered.
Why do I have to sign in before I even know what this does? — User 038/8 overlooked Customize — Reimagine took all the attention.
I thought "Reimagine" was the only option — I didn't even see Customize. — User 066/8 wanted a walkthrough — without one, the AI felt unclear.
I'm in… now what? — User 02Three recommendations to streamline the journey, build trust early, and surface the AI features that make Neighborbrite worth coming back to.
Users hit a login wall before exploring the product — momentum died and early drop-offs spiked.
The signup screen blocked Step 2 — users had to commit before they knew what the tool did.
Move login deeper into the journey — users upload, generate, and customize before they're asked to commit.
Customization was overshadowed by the dominant Reimagine button — 8/8 users walked past the Customize AI tool entirely.
Reimagine was the large, contrasted CTA — Customize was a small link below it.
Uneven weight
Stack Customize and Reimagine vertically with the same visual weight — both AI tools earn a chance to be discovered.
Equal weight
First-time users dropped in with no context — 6/8 said the AI tool felt unclear from the start.
Users landed in a static "How it works" screen with no walkthrough — the first interaction was "figure it out yourself."
A "Watch Demo" entry teaches the AI flow before the user has to try — trust-building before commitment.
New Watch demo entry
By delaying login, balancing the button hierarchy, and onboarding with a demo — Neighborbrite's AI tools finally do the talking.
Designing for an AI-powered product taught me that small decisions early in the journey have outsized impact. Delaying a login. Equalizing two buttons. Adding a demo. Each one tiny on its own — together they unlocked the engagement the platform had been missing.
I also learned to let the product speak before asking the user to commit. When the AI got a chance to demonstrate value first, users opted in willingly. That's a pattern I'll carry into every AI-first product I touch.
Good design for AI isn't about hiding the machinery — it's about making the machinery feel inviting from the very first tap.
