How We Built CivicPass - A Citizenship Exam App

Six months before we shipped CivicPass, the app was a list of flashcards. Swipe right for correct, swipe left for wrong, reset and repeat. It worked - in the sense that you could use it to drill citizenship exam questions. But after testing it ourselves and watching a few people try to prepare for real exams with it, we realised it was solving the wrong problem.

This is the story of what we built instead: why we scrapped the flashcard approach, what we replaced it with, and the things we got wrong along the way that we are still fixing.

The problem we were trying to solve

The citizenship exam prep market has no shortage of resources. There are official government guides, YouTube videos, tutors, community forums, and dozens of apps. So why build another one?

Because almost every existing solution had one of two problems:

• Passive content: Reading guides and watching videos builds familiarity, not exam-ready recall. You feel prepared. You are not.
• Requires internet: Most apps pull questions from a server. If you study on a commute with spotty signal, in a rural area, or in a country where your data plan is expensive, the app becomes unreliable right when you need it most.

We wanted an app that forced active recall - real questions, real feedback, real pressure - and worked completely offline from day one. That constraint turned out to shape every decision we made.

What we tried first (and why it did not work)

The flashcard version had a clean design and a satisfying swipe mechanic. It also had a fundamental problem: users could not tell if they were ready for the exam. They could swipe through 100 questions and feel good, but "feeling good" is not a pass rate. We had no way to tell them - or ourselves - whether the app was actually working.

The second problem was coverage. The French citizenship exam covers five themes: Values, Institutions, Rights, History, and Society. Users who studied our flashcards in random order got strong at some topics and weak at others without realising it, because the random shuffle did not signal which topics needed more attention.

We needed structure. And we needed measurement.

The six study modes we built - and the logic behind them

We replaced the flashcard feed with six distinct modes, each targeting a different stage of preparation:

• Practice Mode: Random questions with instant feedback and explanations. Not just "correct/wrong" - why the answer is what it is. This is the primary study mode for most sessions.
• Theme Practice: Questions filtered by topic, so users can deliberately work on weak areas. If your Institutions score is 60% and everything else is 85%, Theme Practice lets you fix the imbalance.
• Mock Exam: Timed sessions with the exact format of the real exam - 40 questions, 45 minutes for France; 20 questions, 45 minutes for Canada. Official scoring. No help mode during the exam.
• Mistakes Mode: Automatically builds a list of every question the user has answered incorrectly and drills them until they are mastered. This is the feature our testers were most enthusiastic about.
• Progress Tracking: A readiness percentage that updates after every session, broken down by theme. The number has to mean something - we calibrated it so that a user at 85%+ is statistically likely to pass.
• Offline Mode: The entire app - all questions, all modes, all feedback - works without internet. This was not an afterthought. It was a design constraint from day one.

We added one mode at a time, testing each with users before moving on. Mock Exam was the hardest to get right - specifically, the question of what "passing" meant. We went through three different scoring models before settling on the one that matched real exam pass rates accurately enough to be useful as a predictor.

What our beta testers taught us

The most valuable thing we learned during beta testing was not about features - it was about the mental model users bring to exam prep apps.

Most users expected to feel ready before they actually were. They would hit 70% on a few practice sessions and consider themselves prepared. The first version of our readiness tracker reinforced this by showing progress too optimistically. We recalibrated it to be conservative: users now need to consistently score 85%+ across multiple full mock exams before the app tells them they are genuinely ready.

That change made the app feel harder. Users initially pushed back on it. Then several testers went through the real exam and came back to tell us the conservative estimate had been accurate. The exam was harder than they expected it to feel after 70% practice scores. Being told they needed more preparation had been right.

We kept the conservative tracker. Trust over optimism.

What is coming next

CivicPass now covers 8 countries: France, Canada, the United States, the UK, Germany, Spain, Sweden, and Denmark. Each country has 1,000+ mock questions, mock exams in the exact real format, and theme-based progress tracking.

We are also working on a study schedule feature: given your target exam date, CivicPass will generate a daily study plan and track whether you are on pace to be ready. The feature exists in prototype form - we are testing the pacing logic before shipping it.

And we are listening. If you are preparing for a citizenship exam in a country we do not yet cover, contact us at contact@stikastudio.com. Every country request we receive goes into the roadmap queue.

The bottom line

CivicPass started as a flashcard app and shipped as something more structured, more honest about readiness, and more useful for the actual exam. That shift happened because we kept asking whether the app was actually helping people pass - not whether it felt like a good study tool.

There is a difference between those two things. We think it matters.

Stika Studio builds CivicPass and MyRin - mobile apps and games for a global audience.