Building an NCLEX exam prep app from MVP to ExamReady’s flagship product
Role
Platform
Live app
Timeline
purchase conversion
Trial-first paywall: +37% ARPU and +49% trial starts in an A/B test.
D14 retention
Game Mode: paying users who played vs. those who didn’t.
design time for a new app
Design system: down from ~2 weeks, one library for eight apps.
The challenge
NCLEX is a computer-adaptive nursing licensure exam built around different question formats, from single-answer questions to complex, connected clinical case studies.
To stand out among competitors, convert users, and retain them, the app had to be:
Genuinely useful
The exam tests clinical judgment, not memorized facts. Candidates needed practice in every topic and question type, with clear explanations.
Worth practicing daily
There’s too much material to learn in a few sittings. Candidates needed a daily study habit and a reason to keep coming back.
Clear about readiness
Not knowing where you stand is stressful, especially when a retake costs time and money. Candidates needed a clear picture of their progress.
Building the MVP
What we left out
With no user research before launch, we aimed to ship fast and learn from real data. To validate the core practice loop first, we left out a readiness score, a question of the day, clinical case studies, referrals, and a full exam simulation. Several came back in later releases.
Activation
Defining the first aha moments
To activate new users, I chose two moments to show the product’s value: a personal study plan and the first quiz.
From setup to study plan
Five setup questions (exam, date, weak topics, daily time, and when to study) led to a personal study plan: a projected path to readiness and the features to get there. Having invested in their answers, users would feel the app was built for them and be more willing to commit at the paywall.
Research insight:
With a lot of material and limited time, many users didn’t know where to start or how to prepare. They wanted a plan built around their level and schedule.
Success metrics:
Onboarding completion, step drop-off, and paywall-to-trial conversion.
From first quiz to daily practice
I made the Quick 10 Quiz the main action on the Study home screen: exam-style questions with explained answers. Matching the ten-question free limit, it let free users reach the results, with every mistake ready to revisit. I expected a first session that felt like progress to bring them back the next day, and the quiz to become a daily practice for busy days.
Research insight:
For users short on time, even 15–20 minutes had to feel worthwhile. They wanted to see weak spots and felt they were learning when explanations prevented repeat mistakes.
Success metrics:
First-quiz completion rate on install day and D1 retention of users who finished it vs. those who didn’t.
Revenue
Balancing conversion and long-term value
To turn free users into subscribers, I worked on where they met the offer and how it was framed.
Where users met the offer
Free users got ten questions in any mode. In longer quizzes, I proposed a trial offer right after question ten, betting that users partway through a quiz would start a trial to finish it. After the limit, any quiz opened the paywall. It also appeared at the end of onboarding, in banners, and in its own tab.
Success metrics:
Trial starts from the offer after question ten and their trial-to-paid conversion, compared with other entry points.
Leading with the trial
Like most competitors, the MVP paywall offered an annual plan, preselected, and a monthly plan with a free trial. Yet most users closed it without buying either.
Data insight:
The exam dates users entered in onboarding showed they typically had only a few months to prepare. For many, a full year was more than they needed.
We tested a bolder version: only the monthly plan with its free trial up front, and the annual plan one tap away in a bottom sheet. The hypothesis was that one free-to-start option, with nothing to compare, would bring more trial starts and, in the end, more purchases. We expected losing some annual subscriptions to be a small trade-off.
I designed the variant with discount and Limited Time Offer labels.
A/B test · Trial-first paywall
Leading with the trial turned more users into buyers
+49%
trial starts
+43%
purchase conversion
+37%
ARPU
Bringing the annual plan back
The trial-first test won on conversion but carried a long-term cost: with the annual plan out of sight, revenue from it fell by about 80%.
Data insight:
Users stayed about 2.2 months on average, but annual subscribers paid for a full year up front. Even a small rise in the annual share could noticeably lift LTV.
So we brought the annual plan back to the main screen, preselected next to the trial, but not in its MVP form. I made the monthly price the main number on both plans so the annual savings stood out, and added a discount label, expecting more annual sign-ups.
Retention
Giving users reasons to come back
To bring users back, we focused on two things: learning that made every session worth the time, and mechanics that made returning rewarding.
The practice loop
Four quiz modes formed one practice loop: Random Mix for broad practice, Missed Questions to work through mistakes, Saved Questions to revisit what mattered most, and Mastered Questions to reinforce strengths. Each mode worked across all topics or within one.
The Review Questions tab collected every answered question, sorted by result and searchable by keyword. After we saw how often users searched, I added topic filters.
Research insight:
Afraid of forgetting something on exam day, users wanted to work through their weak spots until they’d mastered them.
Success metrics:
Percentage of quizzes taken in each mode, weekly Review use, and D7 retention of users who practice on their own answers vs. only Random Mix.
Seeing progress
The Stats tab showed progress overall and by topic: questions answered and accuracy. I expected the topic breakdown to show users what to practice next and keep them coming back to close the gaps.
Later, we added a readiness score: a single number showing how close users were to being exam-ready, giving them a sense of control.
Research insight:
Without a clear picture of progress, users couldn’t tell which topics needed work or whether they were ready to book the exam, which left them anxious.
Success metrics:
Weekly Stats use, topic quizzes started after viewing Stats, and D30 retention of users who check Stats vs. those who don’t.
Rebuilding the streak
The MVP streak was a simple weekly tracker. Its data revealed a threshold: for paying users without a streak, retention held steady through day five, then dropped sharply, to less than half that of users who reached a five-day streak. Yet only 54% of users who started a streak kept it going the next day, and the average lasted four days, just short of that threshold.
I hypothesized that early milestones and a one-question daily minimum would help more users get past the first few days and build a lasting learning habit.
Research insight:
High-retention learning apps build the daily habit on early wins and a visible next goal, so users always had something within reach.
Streak update · Milestones
More users kept streaks going, and the average more than doubled
MVP → redesign · users who started a streak
54% → 73%
reached a 2-day streak
28% → 49%
reached a 5-day streak
4 → 9 days
average streak length
1.75× more users reached five days, past which retention was over twice as high.
Customer development
Six months after launch, our PM ran an Intercom survey to find users’ top pains. It confirmed the pains from our initial research and showed what users needed next.
The findings moved clinical case studies up the roadmap and shaped how we talked about Cheat Sheets, as a way to ease exam anxiety.
192 respondents, mostly active subscribers · Jul–Sep 2025
What held users back
Top pains, by number of mentions
Exam anxiety
64
Cost
60
Lack of time
46
Hard-to-retain material
45
3.7 times a week
users meant to study but put it off, mostly because of time, fatigue, low motivation, or family
What users needed
Most requested
01
Clinical case studies in the Next Gen NCLEX format
02
More questions
03
A real exam simulation
What helped them study
Next Gen NCLEX case studies
Case studies are the hardest format on the NCLEX: a clinical scenario followed by six connected questions. We expected practice in this format to give users a reason to subscribe and stay.
On a computer, the exam shows the scenario, questions, and patient data side by side. On a phone, they compete for space, and in competitor apps, case studies were hard to follow. I collapsed the scenario above the question and put patient data in a bottom sheet, so users could check it without losing their place.
Research insight:
Case studies were users’ top request in our survey.
Success metrics:
Trials and purchases from free users who tap a case study, completion rate, and subscription renewal vs. users who don’t open them.
Cheat Sheets
Instead of selling cheat sheets separately, we bundled them into the subscription: a cheap standalone pack could have cannibalized it, especially among users who came from cheat sheet ads. We positioned them as quick relief from exam anxiety: short refreshers on the hardest topics.
I turned the content manager’s PDFs into a new category on the Study screen: a row of cards, a full list, and a full-screen viewer with zoom, available offline.
On the paywall, I presented cheat sheets as a gift with the subscription, and a modal after checkout pointed new users straight to them, so they got the promised gift right away. I expected seeing this value early in the trial to help more trials convert.
Research insight:
Exam anxiety was users’ top pain in our survey, and structured cheat sheets were among the study aids they named.
A/B test · Cheat Sheets modal after checkout
Pointing new users to Cheat Sheets turned more trials into subscriptions
Users saw the modal after checkout or didn’t. The gift on the paywall was the same for both groups.
+40%
cheat sheet downloads
+11%
trial-to-paid conversion
+13%
ARPU
The effect held through the following months.
Usage
45%
of paying users opened Cheat Sheets, 8–9 times each on average
Compared with paying users who didn’t open them
Users of Cheat Sheets studied more and stayed longer
Sessions
3.7×
Session length
+23%
Time in the app
4.6×
Quizzes started and finished
1.5–2×
Subscription renewal
31.7% vs. 27.1%
D7 retention
~8×
Game Mode
Before building the game, we tested demand cheaply: a game-style ad creative and a game quiz I designed for our web funnel. Both performed well, so we brought it into the app.
We expected the game to make practice more engaging and bring users back even on days they’d otherwise skip. No competitor had anything like it.
I designed it to feel like a real game: three lives, an endless run of questions, a point for every correct answer, and a personal record to beat, with animations for milestones and new records. It kept the look of the best-performing game ad, so users from that ad or the web funnel already knew it.
Research insight:
Users put off studying 3.7 times a week, mostly because they felt tired or unmotivated. Starting had to feel like a fun challenge, not another task.
Usage
21%
of daily active users played
Compared with paying users who didn’t play
Players stayed longer, and the gap kept growing
D1 retention
D14 retention
Referral
Turning shares into referrals
To add a growth channel beyond paid acquisition, I designed ways for users to share the app, then turned those shares into referrals.
Sharing at the right moments
I mapped the moments when users might want to share: quiz results, cheat sheets, streaks, and Settings. At these moments, I expected users to feel the app helping them and want to pass it on.
Each moment got its own message with an app link, plus a branded image where it fit. Quiz result messages even adapted to the score. Cheat sheets could be shared only as a preview, since full ones came with the subscription. If users took a screenshot of a cheat sheet or streak, the app offered a branded image with a link instead.
Research insight:
Candidates often prepared alongside classmates or colleagues taking the same exam, so they had people to share materials and progress with.
Shares · Dec 2025 – Mar 2026
Cheat sheets led sharing by far
5,148
cheat sheets
506
quiz results
185
from Settings
56
milestones
Cheat sheet shares came from 974 users, about five each.
Inviting friends with a gift
Later, we launched a referral MVP. Every share point became a referral point: each message gained one line, “Get 7 free days of full access with my link — instead of the usual 3.” We expected the longer trial to give friends a reason to install.
I designed two invite points: a Settings banner and a modal after quiz scores of 60%+, when users were likely to feel good about their progress. Both framed the invite as a gift, so sharing felt like helping, not advertising.
For invited friends, an onboarding screen confirmed the gift, and the paywall delivered it: a 7-day trial on both plans.
Data insight:
Users shared cheat sheets far more than their achievements. They passed on what helped them study, and the app itself was exactly that kind of help.
Success metrics:
Banner and modal share rate, invite-to-install rate, and invited users’ onboarding completion, trial starts, and trial-to-paid conversion vs. other new users.
Design system
Scaling to eight apps
After the NCLEX app, my work expanded to the rest of the line.
When we built CDL Prep, the second app, it became clear more exams would follow. It took about two weeks to design in a separate file, and keeping two files in sync took more and more time. So I built a shared design system and moved both apps onto it. The next six apps used it from the start, consistent across the line.
The system included entire screens, adjusted for each exam. When an exam needed something new, I added a component or variant, so nothing lived outside the system.
With tokens, every app came in light and dark themes, on iPhone and iPad. Changing an app’s brand colors, theme, or device took one switch in Figma. I documented the foundations so engineers could rely on the same rules.
Speed and scale
A new app in days, not weeks
design time for a new app
in the design system, from buttons to full screens
Key learnings
01
Watch what a win costs
02
Design for the day users actually have
03
Prove demand before building
04
Check assumptions with users
What I’d do next
Talk to users in depth
Our survey showed what users wanted. In-depth interviews would show why, so we could solve the need behind each request, not just the request. Doing both regularly would keep us building what users find worth paying for.
Validate a full exam simulation
One of users’ top three requests, not built yet because it’s costly. A practice run under real exam conditions could ease exam anxiety, users’ top pain, and show how ready they are. A fake door test could check that demand first.
Measure the referral MVP
I’d compare invited users with those from paid ads on conversion and retention, to see whether referrals are worth investing in. Then I’d A/B test a reward for sharers to check if it pays for itself.
Next case study

























