James Ton | Product Designer (UX/UI)
James Ton | Product Designer (UX/UI)

Vidi

Why new users gave up before their first video

Role
Product designer
Timeline
June 2023
Method
Product analytics and 7 moderated usability tests
Scope of work
Product analytics, usability testing, journey mapping, recommendations
Vidithumbnail

Project summary

Vidi is a SaaS iOS app that helps fashion sellers create video content. Hundreds of templates show you which angles to shoot, and an overlay on the camera helps you line each one up.

The app worked, but new users left before making their first video. I was asked to find out why and how to bring them back. I paired Vidi’s analytics with seven moderated usability tests and mapped how a new user felt through their first session. My recommendation: don’t build the next feature yet. Fix the first mile.

The problem

The app worked as designed, but not as desired

On paper, Vidi did everything right: plenty of templates, a camera that worked, and a finished video in a few steps. The data said new users weren’t getting that far.

38%

picked a template, then never recorded a video

22%

tried more than one template in their first week

3–5days

average time to a first video, in an app designed to be used in minutes

Camera opens

Session length dropped sharply the moment the camera opened

The numbers showed where new users stopped. They couldn’t show why. That’s what the tests were for.

How I framed it

In their first 60 seconds, people are lazy, vain and selfish

I don’t mean that unkindly. It’s the mindset to design for, because a new user’s first session decides whether they come back.

Lazy

No spare time to invest in something they don’t know yet.

For VidiNobody new will study a hundred templates.

Vain

They want to look good using the product, quickly.

For VidiThe first video has to look like something they’d post.

Selfish

They want to know what’s in it for them, right now.

For VidiThe mission can wait. Their product needs to look good today.

The current journey

Following seven people through their first video

I ran moderated tests with seven people and mapped how they felt at each step, from opening the app to trying to keep their video. Tap through the moments below.

Scenario

They’ve just downloaded Vidi and want to test it with a video.

7 people · 5 women, 2 men · aged 22–35
Felt goodFelt bad
1

What got in the way

Five things stood between people and a finished video

Saving meant sharing

5 of 7

The only way to keep a finished video was to share it. People wanted to check it first, or edit it somewhere else before posting. Skipping the share meant losing the video.

“Sharing before I can save is annoying. I don’t want to share this with my network yet”Male, mid 30s
Vidi’s share screen, where the video only saves once it’s shared

Pop-ups on every launch

5 of 7

Pricing, profile set-up and a feature announcement opened every time the app did, before anyone had filmed anything. By the time a helpful tutorial appeared, people skipped it on reflex.

“too many pop ups. I just want to try the app”Male, mid 30s
The Vidi Pro pricing modal that opened on launch

A bug that deleted their work

4 of 7

The app crashed at the editing stage and took everything with it. People had done the hard part, then lost it.

“App broke and I lost all my progress”Male, mid 30s
The editing screen where the app crashed

Editing tools that fought back

Observed in testing

Text and review boxes couldn’t be resized, so they covered the product. Trimming worked some of the time. Simple edits took real effort.

“takes a while to navigate and figure out each feature.”Male, early 20s
The trim toolThe review text box covering the product

A Discovery feed that undersold the app

Observed in testing

Discovery was meant to inspire. Instead it showed blurry, badly lit clips made by other users, which set low expectations before anyone had tried a template.

Four Discovery videos, blurry and poorly lit

The insight

People felt like they were failing before they’d finished the video

While filming, nobody could tell if what they were making was any good. They chased the overlay and doubted the result. Then they saw the finished clip and were pleasantly surprised. Vidi’s best moment came last, and too many people never reached it.

While filming

Unsure, chasing the overlay

“Template quality is much higher but i can’t match it with movement.”
Male, early 20s
“in the clips it looks smooth but i can’t tell while shooting if it’s going to look good.”
Male, early 20s
“I’m unsure if I’m taking a video or a photo”
Female, mid 20s
After seeing the result

Surprised it turned out well

“Video turned out surprisingly well even though I wasn’t convinced while filming”
Male, mid 30s
“final clips actually look pretty good.”
Male, early 20s
“it’s a good video. it doesn’t have to be super fancy for depop”
Female, 26

The decision

Part of the brief was a new feature. I argued for fixing the first mile first

What the brief included

Research a new feature: one-take style videos

Find out what else might help, and add it.

What I recommended

Refine the first mile before building anything new

Tighten the path from sign-up to a first finished video.

Why
01

The core already worked.

People loved the overlay and liked their finished video. The value was there.

02

New users weren’t reaching it.

38% never recorded a video. The blockers sat between them and the payoff.

03

A new feature would inherit the same problems.

Pop-ups, lost videos and fiddly tools would get in its way too. Consolidate first, then build on top.

What I recommended

Get people to their first finished video, fast

Four changes the team could make straight away, then two bigger pieces of work.

Things we could do right now

Pull back the pop-ups

Ask for payment after someone has finished two projects, not on first launch. Review every other modal against the first session.

5 of 7 annoyed

Let people save without sharing

Remove share-to-download. Letting people keep their own video is the simplest way to earn their trust.

5 of 7 confused

Fix the bug that deletes videos

Nothing undoes trust faster than losing the work. This one comes first.

4 of 7 affected

Curate Discovery

Only show approved videos, so the feed shows Vidi at its best, not its worst.

Quick to change
What comes next

Improve onboarding

Get new users to a finished first video as fast as possible, so they feel the payoff before they lose interest.

Review each tool

Shooting, the editing tools and Discovery. Make sure templates are battle-tested and the overlay guides without making people feel they’re failing.

Reflection

What I learned

01

Retention is decided in the first mile.

Everything a new user meets before their first video counts as the product, including pop-ups, save flows and the bugs in between.

02

Fix before you add.

A new feature would have landed in the same broken journey. Consolidating what already worked was the faster path to people coming back.

03

People remember the peak and the end.

Looking back, the peak–end rule explains a lot. Vidi’s best moment came last, and the last moment was often losing the video. I’d design an earlier win and a safer ending.

End of Part I

This is where the review stops

Part I covered the research and what to fix first. Part II, redesigning onboarding and the tools, isn’t part of this case study.

See more work