Case study · Video streaming · Mobile · UX research

Making premium worth paying for.

Memento, a free long-form video app, needed to start earning money without adding to subscription fatigue.

I researched what free users would actually pay for and designed Projects, step-by-step courses from Memento creators. Then I tested with 10 participants and rebuilt the screens, the pricing model and even the feature's name around what they showed me.

Aware: Interests → projects
01AwareInterests → projects
Discover: Projects in the feed
02DiscoverProjects in the feed
Decide: The whole project, no locks
03DecideThe whole project, no locks
Try: 7 days free
04Try7 days free
Start: Straight into Lesson 1
05StartStraight into Lesson 1
Role
UX Researcher & Designer (solo)
Timeline
Oct – Nov 2025
Platform
iOS mobile app
Tools
Figma, Google Forms, FigJam

Overview

Memento is a free mobile and web app for long-form, meaningful video. After two years spent building an engaged audience, the startup wanted to add a paid tier. It set two goals: get new users to subscribe, and convert existing free users, both at sign-in and inside the product.

ProblemAd-free viewing isn't enough reason to pay anymore. Viewers already have too many subscriptions, and they cancel the ones they don't use.
SolutionProjects: step-by-step courses from Memento creators, sitting next to Videos and Ripples. Users hear about it during onboarding and unlock it with one clear offer, a 7-day free trial.
Outcome10 moderated sessions (5 new users, 5 returning) surfaced 4 high-severity issues. They led to 6 redesigns, a switch from a paywall to a 7-day trial, and dropping the "Pointer" sub-brand I loved.

The solution

Here's where it ended up: Projects, introduced at the right moments and unlocked with one 7-day trial. Every step is designed around what the user is thinking, and each one moves a metric I can measure. The rest of this case study shows how I got here.

Learns Projects exists during onboarding, then meets it again right after a related video.

01

Aware

Aware

"Oh, there are courses for the stuff I picked."

Design moveMention Projects right after interests. No price.
Metric it moves% who open a project in week 1
02

Discover

Discover

"That's a course, not a video."

Design moveCourse cards in the feed, shaped differently from videos.
Metric it moves% who tap a course card
03

Interested

Interested

"There's a whole course on this?"

Design moveAfter a related free video, offer the step-by-step version.
Metric it moves% who tap “Turn this into a project”
04

Decide

Decide

"I can see everything I'd get."

Design moveAll lessons, resources and chatroom visible. No locks.
Metric it moves% who tap Start free trial
05

Try

Try

"Free, and they'll remind me. OK."

Design moveOne offer, a reminder before billing, cancel anytime.
Metric it moves% who confirm the trial
06

Stay

Stay

"Let's start Lesson 1."

Design moveStraight into their project, plus a trial timeline.
Metric it moves% who start Lesson 1 on day 1then % still paying after day 7

Research plan

Business objectives
  1. Get new users to subscribe during sign-up
  2. Convert returning free users, both at sign-in and inside the app
Core research question

What makes someone decide a subscription is worth it, and worth keeping?

I broke it into 8 research questions across 3 themes (mapped below).

1Understand the marketSecondary research + competitive / SWOT analysis of YouTube, Vimeo and Twitch
2Hear from the audienceA survey on what people pay for, keep and cancel
3SynthesizePersona + journey map to find where viewers give up
4Validate the design10 moderated usability sessions in 2 separate groups
Questions × methods

Each question had at least two sources, so no single method decided an answer.

Research questionSecondarySurveyCompetitiveUsabilityWhat I found → where it led
Who & how much
Q1Which demographic buys the most subscriptions?18–24, heavy mobile users, became the target group
Q2What's the most people will comfortably pay?Under ~$11/month for entertainment→ Plans · Fix 04
Q3What bundling makes people more likely to subscribe?Discounts for longer plans + cancel anytime→ Plans · Fix 04
What makes it worth it
Q4How do people decide a subscription is worth it?They keep what they use (4 of 5)→ Key insight
Q6What frustrates viewers on other platforms?Endless searching with no clear path→ Journey map
Q7What are viewers' top requested features?Structure: knowing what to watch next→ Projects concept
Market & creators
Q5Which premium features have worked for competitors?All lead with a free trial; "ad-free" alone isn't enough→ Fix 04
Q8What are creators' top requested features?Direct support converts on Twitch → creator credit→ Open: creator interviews
Participants

18–24 year olds who are very comfortable with tech, spend several hours a day on their phones, and for whom video is a big part of daily life. Mixed gender.

Recruiting

Friends and peers, Springboard participants, and same-day sessions in coffee shops to reach people outside my circle.

Honest limits

The survey had 5 responses, so I treated it as directional and always checked it against secondary data. I didn't let it carry a decision alone.

What the research said

People keep paying for what they use. So premium had to be a tool viewers rely on, not just Memento without ads.

Three findings, each backed by more than one source, shaped every design decision that followed.

01

People cancel what they don't use

  • 30%+cancel within 6 months over cost or low valuePrecedence Research, 2025
  • 22%feel streaming isn't worth the moneyMarketWatch, 2025
  • 4 of 5people I surveyed keep a subscription because they use itMy survey · directional

So I designedA premium tool people use, not just "Memento without ads."

02

People want to try before they pay

  • 3 of 3competitors lead with a free trial (7 days to 1 month)My competitive analysis
  • <$11per month felt comfortable; annual discounts and cancel-anytime terms preferredMy survey
  • Tasteof premium itself converts better than opting into a trial laterKoch & Benlian, 2017

So I designedA 7-day free trial, a discounted 3-month plan and a reminder before billing.

03

Everyone sells "ad-free." Nobody sells structure.

  • 2.8%acquisition rate in 2024, down from 4.1% in 2021, so retention is the priorityRecurly, 2025
  • 0 of 3competitors offer guided, project-based learningMy competitive analysis
  • Loopviewers search, find gaps, then search again until they quitJourney map

So I designedProjects: step-by-step courses that give people a reason to stay.

Competitive snapshot

How the market sells premium, and the gap

TwitchYouTube PremiumVimeoMemento opportunity
Entry price$5.99–$11.99/mo$13.99/mo · $7.99 student$12–$75/moUnder $11/mo
Free trial1 month1 month7 days7 days + reminder
What premium sellsCreator support, ad-freeAd-free, background play, downloadsPro tools, privacy, analyticsA learning outcome
The hookwhy people pay & staySupporting a streamer they love: sub badges, emotes, sub-only chatYouTube Music: a huge ad-free library, and often the real reason people payProfessional quality: privacy controls, analytics, team reviewFinishing a real project: lessons, resources and a chatroom in one place
WeaknessSteep learning curve frustrates usersNo feature compelling enough to keep payingBuilt for pros; little for everyday viewersGuided projects, which none offer
View my full competitive analysisFull competitive analysis of Twitch, YouTube and Vimeo

Who I designed for

Viewers weren't short on videos. They were short on a plan.

Persona photo of Huan Chen
Huan Chen, 23Accountant · small city apartment · DIY enthusiast

"I'm completely new to this. What are the ideal starting points?"

Pain points
Information overload Doesn't know where to start
No structure Can't build a learning path
Selection dilemma Which tools and materials work?
View full personaFull persona: Huan Chen
His current journey

Confidence drops at every step, and bottoms out in a loop

01

Search

  • Visits Memento and searches "beginner balcony garden"

“I'm so excited to start growing my own vegetables on my balcony!”

02

Discovery

  • Browses titles, thumbnails and shorts to pick the best tutorial
  • Chooses the one that looks most relevant or beginner-friendly

“I'm not sure which video actually fits my situation.”

03

Extended research

  • Looks for more videos to fill gaps left by the last tutorial
  • Struggles again to choose which one will help

“It's hard to choose which video to watch next.”

04

Continuous loop

Where it breaks
  • Tries to organize the steps but uncovers more gaps, leading back into endless searching

“Every time I try to organize the steps, I find more gaps and end up researching again.”

“Why does this take so long? I just want to get started.”

05

Quit

  • Overloaded by steps, materials and information
  • Decision paralysis leads him to quit

“Hold on… what kind of soil am I supposed to use again?”

“The steps aren't clear and it's too confusing.”

Confidence
ExcitedHopefulUnsureLowest confidenceGives up
The insight

Every tutorial leaves gaps, so he searches again. More videos don't fix the loop. Order and structure do.

→
So I designed

Projects: lessons in order, resources and a chatroom in one place, so the loop turns into a plan.

The concept

Turn related videos into guided, step-by-step projects, with lessons in order plus resources and a discussion room, so viewers stop guessing what to watch next.

I first called it Pointer; the redesign section explains why it became Projects.

The concept

Three ways in, one place to decide, one offer

3 ways in
FeedA row between Videos and Ripples
Search"balcony garden" → a project above the videos
TabBrowse "Pick your next project"
→
1 place to decide
Project pageSee every lesson, the resources and the chatroom before paying
→
1 offer
7-day free trialAsked only when they want to start
Information architecture

What changed after testing

The same map before and after. On After, crossed-out labels show what each piece used to be, so you can see the whole change at once.

Memento app
"New user"First time in the app
Onboarding
Sign upBirthdayPick interests+ "Your interests have projects"
↓
Home feed
Videos"Pointer" row → Projects row
↓
"Build a garden" video
↓
Pointer pop-up"Keep going" promptright after the video
↓
Course pageProject page
Pricing tabs → Lessons · Resources · ChatroomLessons 2–5 locked
↓
Paywall · 48-hr offer7-day free trial
What changedMet cold in a pop-up → introduced during onboarding, then offered again right after a related video.
"Returning user"Already knows Memento
Pointer pop-up on app open"New in Memento" tipdismissible, on the tab
↓
Search "balcony garden"
Pointer row → Project resultVideos
↓
Course pageProject page
Lesson 1 freeLessons 2–5 locked
↓
Paywall7-day free trial
↓
Welcome
Explore Pointer library → Start Lesson 1 + trial timeline
What changed"Did Memento get renamed?" → no new brand to explain, and a week to decide instead of one lesson.
"Projects tab"Part of Memento
Pointer · own logoProjectsMemento style
↓
Explore · My Projects · SavedContinue learning · Topics
↓
Course pageProject page
What changedA second app inside Memento → one more tab, like Ripples.
ScreenSection within screenPop-up / promptOld → newNew or changedRemoved

How it makes money

A feature only works if the business works. Crediting creators raised a question I hadn't planned for: who gets paid, and how?

I didn't model the finances in depth, but I mapped how the money could flow so the design stays realistic. This is a direction to validate, not a final model.

1 · Subscribers pay$10.99/mo

or $26 for 3 months. Student and Family tiers. Most start with the 7-day free trial.

→
2 · Memento keeps most of itThe majority

Covers the app, the team and profit. Premium also includes ad-free and offline viewing, not just Projects.

→
3 · A creator poolSplit by watch time

A share goes into one pool, divided by how much of each creator's Projects people actually watch. Unwatched courses earn nothing.

↺ Creators promote their Projects to their own followers, which brings in new subscribers who wouldn't have come otherwise.
Why not make courses in-house?

Producing courses means paying thousands of dollars up front before anyone subscribes. A shared pool grows with revenue, so the catalog can grow without production costs.

Proven model

YouTube Premium and Skillshare both share subscription revenue with creators based on watch time.

The risk I flagged

At launch the pool is small, so early creators might not earn enough to bother. Possible fix: a guaranteed minimum for launch creators and a small, curated first group.

What testing showed

Participants understood the pricing. What they didn't understand was what they'd be paying for.

10participants, 18–24
5 + 5new users and returning users, recruited separately because in my pilot, people who ran both flows carried what they'd learned into the second one
Does a project shown after a free video create intent to pay?Partly. They didn't know what Pointer was.
Is one free lesson enough to show a course's value?No. They asked for a week.
4 HIGH
3 MED
1 LOW
A

"What is Pointer?"

"I thought the app was called Memento?"

  • HIGHPointer felt like a separate app
  • HIGHCourse cards read as videos
  • LOWOdd spot for a premium tab in the nav

Fixed in 01 · 02 · 06

B

Asked too early, for too little

The paywall felt like friction on "just another video."

  • HIGHPaywall came before the value was clear
  • HIGHOne free lesson wasn't enough to decide
  • MEDPrice was clear; worth wasn't

Fixed in 03 · 04

C

The value disappeared after paying

"Start Learning" sent people away from their lesson.

  • MEDWelcome screen didn't confirm the value
  • MEDNo one found Resources or the Chatroom

Fixed in 03 · 05

Severity key
HIGHMultiple users couldn't complete the task or struggled
MEDCompleted, but raised concerns
LOWCosmetic or minor

Moderated think-aloud sessions on grayscale wireframes.

The hard call

The name I loved was the problem.

Introducing Pointer, its own brand
1 · What I wanted

"Pointer"

Like the dog breed, it points you where to go. I gave it its own name, logo and colors.

"I thought the app was called Memento?"

"Did Memento get renamed to Pointer?"

2 · What testing showed

"A different app"

Confusion in both rounds. Adding "Premium Feature" labels in v1 didn't fix it.

Before: Pointer
→
After: Projects
3 · What I decided

"Projects"

The cause was a second brand inside an app. A plain word beside Videos and Ripples needs no explaining.

The lesson: users pay for clarity, not my symbolism. Cutting what I loved because the evidence said so was the biggest decision in this project.

Findings → redesign

Each fix shows the wireframe participants tested, my first hi-fi (v1), and the second round (v2). Yellow callouts point to exactly what changed on each screen.

HIGHFIX 01 · "I THOUGHT THE APP WAS CALLED MEMENTO?"
Theme A "What is Pointer?" · fixes 01 02 06↑ All themes

From "Pointer" to Projects

What happened in testing

Participants didn't understand what Pointer was, or that it belonged to Memento. The confusion followed them from the pop-up to the paywall, the welcome screen and the nav tab. It came up in both rounds.

Why this change

The finding was a symptom. The cause was structural: Pointer had its own logo, gradient and wordmark, so tapping into it felt like leaving the app. v1 added "Premium Feature" labels and didn't fix it. v2 removes the sub-brand completely.

Tested
Tested pop-up
"Pointer" with no link to Memento → users thought it was another app
"Our premium feature," but whose? Users couldn't tell who was selling it
Hi-fi v1
Hi-fi v1 Pointer tab
Pointer's own logo and colors replace Memento's → "Did I leave the app?"
A "Pointer" tab with a dog icon reads as a second app, not a Memento feature
  • Labels didn't help: users still asked "I thought this was Memento?"
Hi-fi v2
Hi-fi v2 Projects tab
Progress first gives subscribers a reason to come back; the offer stays out of the way
Same header, colors and type as Memento → no "is this another app?" doubt
Same cards everywhere → users recognize a course instantly
  • Why it sells: people don't pay for things they don't understand. Making Projects feel native removes the "is this another app?" doubt that stalled the trial decision.
The hard callWhy I let go of a name I loved → read the story
Returning users

"Did Memento get renamed to Pointer?"

Returning users already know Memento. In v1 I tried to fix the confusion with a better introduction: a full-screen "Introducing Pointer · Premium Feature" pop-up. Users still read Pointer as its own brand, and a new logo and gradient taking over the screen looked like a rebrand. That's when I renamed it. In v2 there's nothing new to explain, just a small, dismissible "New in Memento" tip on the Projects tab.

Tested · Returning
Tested pop-up on app open
Blocks the app before users do anything → closed without reading
A new name in a pop-up made returning users ask "did Memento rename?"
Hi-fi v1 · Returning
Introducing Pointer premium feature
A nicer pop-up, but still interrupts people who opened the app to watch
New logo + new name → still read as a rebrand; the confusion stayed
Hi-fi v2 · Returning
New in Memento tip pointing at the Projects tab
A "New" dot users choose to tap → invites them instead of interrupting
Creators they already trust → a familiar reason to try Premium
"Projects" needs no explaining → no rebrand confusion for loyal users
  • Why it sells: returning users are the cheapest people to convert because they already trust Memento. A tip they choose to open keeps that trust instead of interrupting it.
HIGHFIX 02
Theme A "What is Pointer?" · fixes 01 02 06↑ All themes

Make a course look like a course

What happened in testing

Course cards used the same gray box and play icon as every video. Participants read the row as recommendations and scrolled right past the core feature.

Why this change

Projects, Videos and Ripples are three kinds of content, so they need three different shapes. In v1, tall portrait cards still read as Ripples (our Shorts). In v2, the shape says "several lessons" before anyone reads the title.

Tested
Tested search
No lesson count → users didn't know it was a course
Looks like every other video → users scrolled right past the core feature
Hi-fi v1
Hi-fi v1 home
Better covers, but still no sign of multiple lessons
Same shape as Ripples → mistaken for short videos, not courses
Hi-fi v2
Hi-fi v2 home
Sits beside Videos and Ripples → clearly part of Memento
Stacked edges = several lessons → recognized as a course at a glance, so it gets the tap
A creator they already follow → trust before they even tap
  • Why it sells: a feature people scroll past can't convert anyone. Cards that read as courses get the tap, and the tap is the first step toward a trial.
Card states: not started vs 40% complete

The card's only job is the tap. Tapping opens the project page, where people see everything and decide. So the card never sells; it just reads as a course.

  • Not started: empty pills + "5 lessons · 2 hrs"
  • In progress: pills fill in, "40% complete"
  • The offer appears once: "Free for 7 days" in the row header
HIGHMEDFIX 03
Theme B Asked too early · fixes 03 04Theme C Value disappeared after paying · fixes 03 05↑ All themes

Sell the whole project, not the next video

What happened in testing

After watching Lesson 1, most participants weren't ready to pay for Lesson 2: "I'd need to see more to know if it's worth it." One video didn't convince them. And nobody noticed Resources or the Chatroom, the perks that make a project more than a playlist.

Why this change

With the 7-day trial, nothing on this page needs a lock. Instead of putting a paywall on every lesson, v2 shows what you get: 5 lessons, Resources and the Chatroom as three clear perks, with one sticky "Start 7-Day Free Trial" button. People decide on the whole project, not on one video.

Tested
Tested course page
Pricing where lessons should be → felt like a sales page, not a course
Locks after one video → "why pay for just another video?"
Hi-fi v1
Hi-fi v1 course page
Resources and Chatroom were hidden in tabs → nobody found the perks worth paying for
A paywall on every lesson → one free video wasn't enough to convince users
Hi-fi v2
Hi-fi v2 project page
Perks shown up front → users see the full value before deciding
No locks → users judge the whole project, not one video
One clear button + "No charge today" → removes the fear of committing
  • Why it sells: people commit to an outcome ("grow my garden"), not to one video. Showing the whole project and its perks gives them a reason to start a trial instead of a reason to leave.
HIGHMEDFIX 04
Theme B Asked too early · fixes 03 04↑ All themes

From "1 free lesson, then pay" to one 7-day trial

What happened in testing

Participants understood the plans but felt pressured by "48 Hour Offer." One free lesson wasn't enough to judge a course; they asked for a week.

Why this change

The value of structure only shows up over several lessons. My competitive analysis matched what participants said: Twitch, YouTube and Vimeo all lead with a trial. So I dropped the free-lesson model for one clear offer. And because this audience is tired of subscriptions, the reminder before the trial ends answers their biggest fear.

Tested
Tested paywall
Interrupts mid-course → felt like being held hostage
Fake urgency → users felt pressured, not persuaded
Hi-fi v1
Hi-fi v1 paywall
Tied to the course, but still asks for money after one video
One lesson isn't enough to judge a whole course → people hesitated
Hi-fi v2
Hi-fi v2 trial
A full free week → enough time to see the course is worth it
A reminder before billing → answers the #1 fear: surprise charges
No risk today → easier to say yes
  • Why it sells: a free week removes the risk of paying first, and the reminder removes the fear of a surprise charge. Both tackle the top reasons this audience avoids subscriptions.
MEDFIX 05
Theme C Value disappeared after paying · fixes 03 05↑ All themes

Confirm the value, then start learning

What happened in testing

The welcome screen was a paragraph of text, and the CTA sent people away from the course they'd chosen.

Why this change

The moment after signing up should confirm "you made the right call" and get users straight back to the project they picked. v1 still had the full-screen Pointer gradient, the look we'd removed everywhere else. v2 is Memento-native and answers this audience's biggest worry, forgetting to cancel, with a simple timeline.

Tested
Tested welcome
Nobody reads a wall of text → no sense of what they just bought
Sends users away from their course → breaks their momentum
Hi-fi v1
Hi-fi v1 welcome
Benefits are there, but repeat the same line, so they say nothing
Pointer's gradient again → "am I still in Memento?"
Hi-fi v2
Hi-fi v2 welcome
Memento's look → users know exactly where they are
Their own project, ready to go → straight back into learning
A clear timeline → no surprise charges builds trust that keeps subscribers
3 short lines → the value is understood in seconds
  • Why it sells: starting a trial isn't the win; keeping the subscriber after day 7 is. Getting people into Lesson 1 right away gets them using it, and the research showed people keep what they use.
HIGHFIX 06 · NEW IN V2
Theme A "What is Pointer?" · fixes 01 02 06↑ All themes

Introduce Projects in steps, not all at once

What happened in testing

Users either met the feature with no context, or hit a paywall before they understood it.

Why this change

Research said not to sell on day one, but nothing said premium had to stay hidden. Choosing interests is the one moment a new user tells us what they want to learn, so pointing out that those interests have guided projects is useful information, not a pitch.

Aware · onboarding
Onboarding step
Uses what they just told us they love → curiosity before any ask
No prices, no pressure → plants the idea of Premium without scaring new users off
  • Why it sells: it uses the interests people just picked ("you could grow your own veggies") to plant the idea of Premium at no cost to them. They meet Projects already curious, not cold.
Interested · in context
After a related video
Appears the moment they finish a related video, when interest peaks
The actual lesson plan → they see the path, not a sales pitch
One offer, easy to dismiss → "Not now" keeps trust intact
  • Why it sells: they just watched a free video on this exact topic, so interest is at its peak. Offering the structured version right then turns a passing interest into a reason to try Premium.
Ask · at intent
Trial screen
One simple offer → nothing to compare or second-guess
  • Asked only once they open a project, when they're already interested
  • Why it sells: asking only when someone opens a project means the offer meets intent. Fewer, better-timed asks avoid the upsell fatigue that makes people ignore offers.

How I'd measure success

This was a concept, so there's no live data yet. Here's the measurement plan I'd hand to the team, with each metric tied to the fix it tests.

The funnel I'd instrument

Where each fix should show up

Sees Projectsonboarding · feed · tab
Fix 01, 06
Opens a projectcard tap rate
Fix 02
Starts 7-day trialtrial start rate
Fix 03, 04
Pays after day 7trial → paid
Fix 04, 05
Still subscribed at day 90retention
Resources · Chatroom

Bar lengths show the expected shape of the funnel, not real results. Each step's baseline would be set from the first two weeks of live data.

Hypothesis to validate

Do the perks keep people subscribed?

0%50%100%Day 0Day 30Day 60Day 90 Uses Resources or Chatroom, day 7Uses Resources or Chatroom, day 14Uses Resources or Chatroom, day 30Uses Resources or Chatroom, day 60Uses Resources or Chatroom, day 90Lessons only, day 7Lessons only, day 14Lessons only, day 30Lessons only, day 60Lessons only, day 90 Uses Resourcesor Chatroom Lessons only

Dotted lines mark the day 5 reminder and day 7 billing. Illustrative curves for the hypothesis I'd test: subscribers who use Resources or Chatroom stay subscribed longer. If it holds, those perks deserve more space before the ask.

Re-test the rename

Do people now see Projects as part of Memento?

In testing, people thought Pointer was a separate app. To check the rename fixed that, I'd show 5 new people the Projects screen for 5 seconds, then ask: "What app is this?"

Goal: at least 4 of 5 answer "Memento." This is a target, not a result.

A/B test

Does the onboarding step pay off?

ASees "Your interests have projects"
vs
BSkips straight to the feed

Measure: % who open a project in week 1. If A doesn't beat B, cut the screen and keep onboarding short.

What I'd do next

Interview creators

Test whether a watch-time revenue share is worth their time. This answers research question 8 on creator needs.

Test Projects as the center tab

The one LOW finding: does a more prominent spot raise discovery without hurting the core video experience?

Add the annual plan

Participants and the survey both asked for a yearly discount next to monthly and 3-month billing.

"Continue learning" on Home

A "Lesson 2 of 5" card at the top of the feed: the strongest everyday reminder of what Premium gives them.

Reusable test kit: the two task scripts, the 4-level severity scale and the findings template are written so any designer or PM can run round 2 without me.

The design system underneath

Memento runs on tokens in two layers: raw color ramps, then named roles that screens actually use. Every redesign in this case study was built from them, so Projects looks like Memento and not a separate app.

Color · 2 layers

Primitives feed roles

Blue
Pink
Dark Blue
White

4 of 6 ramps · 50 to 900

  • Surface/Primary→ blue-500 · Buttons, active tab
  • Text/Heading→ dark-blue-700 · Titles and body
  • Surface/Page→ white-50 · App background
  • Border/Secondary→ pink-500 · Progress, highlights
  • Logo/Primary→ pink-500 · Brand mark
  • Icon/Disabled→ dark-blue-200 · Inactive icons
Type · SF Pro

One family, all on a 4-point grid

Title / H148/56Projects
Title / H332/40Projects
Title / H424/32Structured Projects
Title / H520/28Structured Projects
Body16/24Structured Projects
Caption12/16Structured Projects
RegularMediumSemiboldBold
Components

The pieces Projects is built from

Padding and gaps in 4s and 8s · radius 4 / 8 / 16 / full

Beginner5 lessonsGardeningBaking
40% complete
Project card component in its not-started and in-progress states
4 · tags8 · cards16 · sheetsfull · pills

Reflection

Wording and visuals have to be designed together

"Lesson" meant nothing while the card still looked like a video. The UI has to say "course" before the copy does.

When you ask matters as much as how

The same paywall feels like friction or like an invitation depending on whether the user has felt the value first.

Structure beats labels

When users are confused, adding explanation rarely fixes it. Changing what the screen is does.