How I made a car quiz that admits there is no single right answer

Paul — Mon, 14 Sep 2026 07:04:21 +0000

Most car quizzes are really personality quizzes with a car-shaped result at the end.

I wanted to build one that felt a little more honest. Not because an algorithm can know the perfect car for someone, but because the process of choosing a car usually has competing constraints: budget, passengers, daily driving, weekend plans, tolerance for inconvenience, and the kind of compromises a person is actually willing to make.

That became RIDE OR DIE, a car-matching quiz with a roasting AI and buy links for the final recommendations.

The feature is the difficulty selector

The first decision I made was to let people choose how much effort they want to spend.

Easy is a short warm-up with 5–6 questions. Medium goes to 10–12 questions. Hard can use up to 30 questions for people who want the recommendation to dig deeper.

This sounds like a small interface choice, but it changes the whole experience. A quiz that assumes everyone wants the maximum number of questions is asking for commitment before it has earned attention. A quiz that is too short can feel random. Giving people control lets them decide whether they want a quick nudge or a more considered match.

The important part is that the question bank is larger than any one quiz. There are 100+ questions available in total, but a user does not see all of them. The selected difficulty controls how much evidence the matching flow collects.

Why the roast exists

A serious recommendation can become very dry, very quickly. Make, model, drivetrain, price range — useful, but forgettable.

The roast is there to give the result a personality. It should feel like a friend pointing out the contradiction between “I need something practical” and the sports car tab that has been open for three weeks. The joke is not the recommendation; it is the wrapper that makes the recommendation easier to remember.

That distinction matters. The product still needs to respect the answers and the budget. The AI can be playful without pretending that every choice is equally sensible.

The result has to be actionable

A quiz result that says “you are a hatchback person” is not enough. RIDE OR DIE pairs the match with links where someone can continue researching or look for a real purchase.

That turns the result from a personality label into a next step. The quiz helps narrow the search, the explanation makes the match feel understandable, and the links give the user somewhere to go afterward.

I am still tuning the balance between useful detail and a fast first experience. That is the part I like about shipping small products: a feature can be technically finished while the experience is still teaching you what it wants to become.

Try RIDE OR DIE here: https://appbioniclabs.s.gy/agfYHN

I also build VibeCard and AnafTracker. You can find the rest of my apps at https://appbioniclabs.xyz.

I built a digital business card that tells you who actually looked at it

Paul — Sun, 13 Sep 2026 07:59:00 +0000

Paper business cards die quietly. You hand one over at a meetup, it survives the evening in a pocket, and then it's gone — laundry, drawer, trash. Even when it survives, you never know if the other person actually did anything with it.

The digital tools that replaced them have the opposite problem. Link-in-bio pages are pretty, but they're blind: you share a link, someone glances at it, and you have no idea who they were, when they looked, or whether they cared.

So I built VibeCard — a digital business card with lead tracking built in.

The idea in one sentence

Every time someone opens your card, that shouldn't just be a "view". It should be a lead.

Here's the flow:

Card shared — your card goes live via QR code, link, or NFC tap (works on any phone, no app needed)
Lead captured — each scan auto-creates a tagged lead in your dashboard
Follow up — you get a reminder before the prospect goes cold

That's it. The card is the bait; the dashboard is the point.

What's in it

25+ platform integrations — social, messaging, payments, booking, all in one card
12 templates with profile photo
Follow-up reminders so no lead rots
Team plan — assign leads and follow-ups to teammates
Analytics & CSV export — see which channels actually convert
GDPR-compliant, built in Romania 🇷🇴

I deliberately kept the solo plan free forever — one account, no credit card, no design fees. If you're a freelancer or consultant, you can go full workflow on $0. The Business plan (team features, advanced analytics) is $15 during the early launch and locks in forever — it goes to $39 after.

The honest build story

I vibe-code my apps on Base44 and ship them fast, then let real usage tell me what to fix. VibeCard started as exactly what everyone builds first: a pretty card with links. Then I looked at the data and saw the thing nobody was building: the card was getting opened, but the maker never found out who. That gap between "shared" and "followed up" is where deals die — so V2.0 is basically a mini-CRM wearing a business card costume.

What I'd love from you

I'm building this in public, so feedback beats applause:

Try the free card: vibecard.digital
Roast the landing page
Tell me what a lead-capture flow is missing for YOUR workflow

I also make RIDE OR DIE (a car-matching quiz) and AnafTracker (Romanian tax deadline tracking) — you can see everything at appbioniclabs.xyz.

If you want to see how it works, scan any QR on the site with your phone — you'll get the exact experience your prospects get.

DEV Community: Paul