Case Study · Dyne Technologies

Meetup Creation

Role UX DesignerCompany Dyne TechnologiesTimeline 1 monthTeam 1 PM, 2 iOS engineers, 1 QA

-23%user drop-off

The problem: Creating a meetup is Dyne’s core action and it was where users gave up — new users overloaded by unfamiliar content, returning users forced to restart over a coupon they forgot to buy.
The call: I tested the cheap fixes first and let them fail on their own merits, which is what made the case for restructuring the flow into modular steps.
The outcome: Drop-off fell 23% within three months and CSAT rose 62% among surveyed users.

Dyne's meetup creation feature — The meetup creation feature

The problem

Dyne is built around getting people together over a meal, so creating a meetup is the app's core action. It was also where users gave up. Drop-off was high and adoption was low, and testing showed the flow was failing two different groups for two different reasons.

New users hit information overload. Too much unfamiliar content was front-loaded into the first steps, before anyone had a reason to care about it.

Returning users hit a wall. Meetups required a pre-purchased coupon. People who forgot to buy one first had to abandon the flow and start over from the beginning, having already done the work once.

Annotated problems with the old meetup feature — Complaints on the old flow, prioritised by volumeClick to enlarge

Research

I ran usability sessions with both new and returning users and mapped the journey for each segment, since the friction points weren't in the same places. The number that settled the argument: seven of eight participants left the flow exhausted.

Not confused, not blocked. Exhausted. That pointed at the shape of the whole flow rather than any individual screen, which mattered for what came next.

User journey map covering new and returning users — Journey map: a new user and a returning user, both exhausted by the endClick to enlarge

What didn’t work

Trying the cheap fixes first

I didn't open with a rebuild. A structural redesign is expensive and easy to argue against, so I tested the low-effort options first and let them fail on their own merits:

Interactive tooltips to guide first-time users through the unfamiliar parts
In-flow coupon purchase, so forgetting one didn't mean starting over

These helped at the margins but didn't move the core problem, which was that the flow asked for too much at once regardless of how well each step was explained. Having the evidence that patches weren't enough is what made the case for restructuring.

Failure points in the coupon marketplace — In-flow coupon purchase, one of the low-effort fixes

The redesign

I looked at Instagram's post creation flow as a model, not for its visual design but for how it handles the same underlying problem: a multi-step creation task that never feels like one. Three principles carried over.

Bite-sized steps. Break the process into pieces small enough that none of them reads as work.
Integrate related features. Bring the thing you need into the moment you need it, rather than making it a prerequisite.
Group related actions. Cluster decisions that belong together so users switch context less.

The modularized flow addressed both segments at once. Coupon options were scoped to the selected restaurant, which cut the choice down to what was actually relevant and folded the purchase into the flow instead of before it. Related actions were grouped into digestible stages, so new users met unfamiliar content a little at a time.

Patched Rebuilt

One-time tooltips guiding first-run users through the original flow — One-time tooltips explaining the existing steps

The redesigned modular meetup creation flow — The same task, broken into modular stages

The patch explained the flow better. The rebuild meant there was less to explain.

Instagram's post creation flow — Instagram, used as a structural reference

The result

-23%

drop-off within 3 months

+62%

CSAT among surveyed users

$1.6M in seed funding — Dyne’s $1.6M seed round, announced March 2023

Want more detail on this one?

This page is the short version. I’m happy to walk through the earlier fixes we tried, the usability sessions, or anything else that didn’t make the cut.

Get in touch ↗