Haven Onboarding: Designing for Trust & Belonging

Back to Projects

Category:

Mobile Design

Client:

Haven

Redesigned onboarding to improve new user activation and community engagement

My Role

UX & Interaction Design

Team

Rommia White (CEO), Rajvi Sanghvi (UX Designer), Nnameka (Front-End Developer), Amar (QA Engineer)

Duration

2 Months From Sept'25 to Oct'25

Methods

Rapid Prototyping, High-Fidelity Prototyping, Usability Testing, User Interviews

OVERVIEW

About Haven

Haven is a peer-to-peer community platform designed to make it easier for people to give and receive help within their local circles. Think of it as a "karma economy" where:

S
h
a
r
e

Earn points by giving.
More points = higher karma.

S
h
o
w

U
p

Answer when called. Decline to help your haven more than 3x in a row, and risk suspension from the commune.

M
a
k
e

F
r
i
e
n
d
s

Every successful exchange creates exponential positivity. Be kind, do your part, and watch your world expand.

The Challenge

Too many verification steps before users could see any value

Felt confused about what to do after verification

RESEARCH

Survey Results

n=27 across 2 apartment buildings (Live for MVP testing)

Why people join:

78% joined to meet their neighbors and not to save money or get help

37% were "just curious", suggesting low-commitment entry matters

Whats's blocking them:

48% said "I'm not sure what to post"

30% said "I'm not sure who else is using Haven or if I trust them"

7% didn't know what a "spark" was

What we learned

Users didn't mind the steps but they minded not knowing why

Verification must be clearly justified, not just speed

Users compared Haven to Facebook Groups, WhatsApp, and Nextdoor

Show the community is alive before asking for commitment

Trust signals matter: real photos, real names

Profile setup isn't bureaucracy, it's community building

Users felt hesitant asking for help

Tone must feel supportive and permission-giving

Non-Negotiable Constraints

Invite Only Access

Haven communities are built on real relationships. Invite-only access ensures every member is vouched for by someone already in the network.

Manual Approval by Admins

Each request is reviewed by a human community admin and not an algorithm. This preserves neighborhood-level trust and accountability.

Phone + email verification must both occur

Both details are required so community admins can verify a user's identity and confirm they actually live in the building or neighborhood they're requesting to join.

These constraints were required to preserve safety and community trust

DESIGN CHALLENGE

How might we redesign Haven’s onboarding experience to

W
h
e
n

k
a
r
m
a

i
s

y
o
u
r

c
u
r
r
e
n
c
y

a
n
d

e
v
e
r
y
t
h
i
n
g

i
s

f
r
e
e

s
i
m
p
l
i
f
y

e
n
t
r
y

SOLUTION

Designing the Flow

Streamlining Verification

Phone Verification

Email Verification

Building Trust Through Profiles

complete your profile

it builds trust and makes sharing easier. sparks with profiles receive more responses.

9:41

bio

help your haven get to know you in a few words.

social links

save and continue

skip

add photo

old design

9:41

Continue

First name

John

Last name

Doe

Having trouble? Contact Support

Haven runs on trust

A name and a photo helps your community know who they’re sharing with.

9:41

Bio

Unlicensed therapist, dating profile curator, hoarder of endless reusable jars.

Having trouble? Contact Support

Finally...

Share a few words about you.
Members with bios earn more karma.

Continue

Rewarding

Explaining Points/sparks system

share actions, not words.

use sparks to take actions in the app. earn points for each contribution, unlock rewards, and build karma along the way.

Skip

9:41

old design

Haven Entry

joining an existing haven?

if so, you may have received a code. enter it here so we can match you to the right one.

dupont

submit

skip

already submitted a code?

space

123

9:41

old design

Explore

My Sparks

Messages

Find your Haven

Choose your building or neighborhood to
start exploring

Morning Heights, San Francisco, CA

Strata Heights

Request to Join

Request Submitted + Entry (Enable Notifications)

Explore

My Sparks

Messages

Recent Favor

Helped debug Jordan's website

New York

Matthew L.

Jordan J.

Recent Favor

Fed Tiffany's cat while she was

away on a work trip

Fed Tiffany's cat while she was

away on a work trip

New York

Tif W.

Recent Offer

Taught an Intro to Boxing class at

her apartment complex

Taught an Intro to Boxing class at

her apartment complex

New York

Anna S.

What I Learned:

Building emotional alignment BEFORE data collection is crucial for community products. The welcome animation wasn't "nice to have", it was the foundation of trust.

What I'd Do Differently

Test different animation styles earlier. We spent time on copy when illustration style might have mattered more.

Test Early, Test Often

What I Learned:

Points only mattered when contextualized (e.g., "points = higher trust = faster help"). Abstract gamification fails; connected gamification succeeds.

What I'd Do Differently

Prototype 3 different incentive models (points, badges, social proof) and test with users before committing to one.

Design for Belonging

What I Learned:

Onboarding isn't a hurdle to overcome; it's the first act of participation. Framing each step as "joining a community" vs "completing a task" changed user perception entirely.

What I'd Do Differently

Interview more community organizers (offline) to understand belonging psychology before designing.

›

Priya Jain

Priya Jain

Haven Onboarding: Designing for Trust & Belonging

Redesigned onboarding to improve new user activation and community engagement

My Role

Team

Duration

Methods

About Haven

The Challenge

Survey Results

What we learned

Verification must be clearly justified, not just speed

Show the community is alive before asking for commitment

Profile setup isn't bureaucracy, it's community building

Tone must feel supportive and permission-giving

Non-Negotiable Constraints

These constraints were required to preserve safety and community trust

Designing the Flow

Streamlining Verification

Building Trust Through Profiles

Rewarding

Explaining Points/sparks system

Haven Entry

Request Submitted + Entry (Enable Notifications)

What I Learned:

Building emotional alignment BEFORE data collection is crucial for community products. The welcome animation wasn't "nice to have", it was the foundation of trust.

What I'd Do Differently

Test different animation styles earlier. We spent time on copy when illustration style might have mattered more.

What I Learned:

Points only mattered when contextualized (e.g., "points = higher trust = faster help"). Abstract gamification fails; connected gamification succeeds.

What I'd Do Differently

Prototype 3 different incentive models (points, badges, social proof) and test with users before committing to one.

What I Learned:

Onboarding isn't a hurdle to overcome; it's the first act of participation. Framing each step as "joining a community" vs "completing a task" changed user perception entirely.

What I'd Do Differently

Interview more community organizers (offline) to understand belonging psychology before designing.

Redesigned onboarding to improve new user activation and community engagement

My Role

Team

Duration

Methods

About Haven

The Challenge

Survey Results

What we learned

Verification must be clearly justified, not just speed

Show the community is alive before asking for commitment

Profile setup isn't bureaucracy, it's community building

Tone must feel supportive and permission-giving

Non-Negotiable Constraints

These constraints were required to preserve safety and community trust

Designing the Flow

Streamlining Verification

Building Trust Through Profiles

Rewarding

Explaining Points/sparks system

Haven Entry

Request Submitted + Entry (Enable Notifications)

What I Learned: Building emotional alignment BEFORE data collection is crucial for community products. The welcome animation wasn't "nice to have", it was the foundation of trust.

What I'd Do DifferentlyTest different animation styles earlier. We spent time on copy when illustration style might have mattered more.

What I Learned: Points only mattered when contextualized (e.g., "points = higher trust = faster help"). Abstract gamification fails; connected gamification succeeds.

What I'd Do DifferentlyPrototype 3 different incentive models (points, badges, social proof) and test with users before committing to one.

What I Learned: Onboarding isn't a hurdle to overcome; it's the first act of participation. Framing each step as "joining a community" vs "completing a task" changed user perception entirely.

What I'd Do DifferentlyInterview more community organizers (offline) to understand belonging psychology before designing.

What I Learned:

Building emotional alignment BEFORE data collection is crucial for community products. The welcome animation wasn't "nice to have", it was the foundation of trust.

What I'd Do Differently

Test different animation styles earlier. We spent time on copy when illustration style might have mattered more.

What I Learned:

Points only mattered when contextualized (e.g., "points = higher trust = faster help"). Abstract gamification fails; connected gamification succeeds.

What I'd Do Differently

Prototype 3 different incentive models (points, badges, social proof) and test with users before committing to one.

What I Learned:

Onboarding isn't a hurdle to overcome; it's the first act of participation. Framing each step as "joining a community" vs "completing a task" changed user perception entirely.

What I'd Do Differently

Interview more community organizers (offline) to understand belonging psychology before designing.