$0.99 to see results, then 3 days later it charges 39.99: Cognius makes its money from a quiz page, the app just delivers the results

Cognius: Know Yourself is an "AI psychology self-test" app that launched at the end of September. Over the past two weeks it has charted in health categories in 11 countries including Korea, Czechia and Egypt. But App Store isn't what pulled it up. In front of it sits a web quiz page that has been running ads since June. It first takes 0.99 dollars to show test results, then if you don't cancel it charges 3 dollars a month after 39.99 days, and only after you pay does it ask you to download the app. The app's downloads and ratings are really carried along by the sales that quiz page makes.

A Korean one-star review

In 2026 Nov 3, a 1-star review appeared in the Korean App Store:

I clicked through an Instagram ad to a Korean page that looked good. At about 0.9 dollars it seemed cheap, so I paid. Then 3 days later it auto-charged me 39 dollars. The app is all in English and no help at all, I don't even know what this app is, so why was I charged?

The same week, the Korean store had a dozen more reviews with similar titles: "Auto-billing", "Charged without consent", "50k won gone, refund me". The Korean store has 20 ratings, averaging 1.7 stars (App Store Korean store, 11-06). A US store review from Oct 30 was clearer: "I finished the test, paid 1 dollars with Apple Pay to see my result. After paying, it asked me to download the app to see the result I had already paid for."

Put these reviews together and you have Cognius's entire business.

What it is

On the store page it's an "AI emotional companion": 40 psychology mini-tests (attachment style, stress response, communication style, empathy, plus ADHD and autism self-tests), with AI turning results into card-style readings and an AI guide that remembers your results and chats with you. The developer is Mindly Apps Ltd, registered in Nicosia, Cyprus, and it has only this one app. It supports 11 languages, and the store description is written in Spanish.

Its real entry point is the web: quiz.cognius.app, titled "What kind of person are you?", starts by asking for your gender. cognius.app/iqtest is the IQ test. The site is built on an off-the-shelf quiz-funnel tool, runs dozens of page variants for A/B testing, and takes payment through Stripe, bypassing Apple.

How much it grew

The app numbers are small. The countries where it charted, like Czechia, Slovakia, Azerbaijan and Kuwait, have very low bars in health charts, where a few downloads can reach the top 200. The real sign of scale is 876 reviews on Trustpilot: for an app only a month old, the web side has likely been paid for by thousands or tens of thousands of people (inferred, roughly estimated from review count).

The timeline shows the order: the Trustpilot page was created in June, and by Jul 11 there was already a Facebook post saying "Cognius app is a scam" with 141 comments, all about the 39.99 dollar charge. The web quiz page ran at least three months before the app.

Where the money comes from

Pricing is in cognius.app's subscription terms:

• Trial 0.99 dollars (or 0.99 euros, 0.99 pounds) for 3 days;
• Auto-renews monthly at 39.99 dollars (39.99 euros, 34.99 pounds) unless canceled;
• There are also 7-day 4.79-dollar and 30-day 7.99-dollar entry prices, which likewise convert to 39.99 dollars a month when they expire;
• Subscriptions bought on the web can only be canceled in the web account or by contacting support.

In this setup, the 0.99 dollars isn't there to make money, it's there to get your card on file. The real money is the 39.99-dollar charge on day 4. A US-store user wrote in Oct 23: a 39.99-dollar charge failed twice, then a 27.99-dollar attempt also failed, and finally a 20-dollar charge went through. Looks like they retry at a lower price after a failed charge (my inference, based on this one review only).

It also built a separate page, iq.cognius.app, titled "Questions About Your Cognius Charge?", which says "Skip the 30-day bank dispute process and contact us directly, we usually resolve it within hours." Almost every negative Trustpilot review gets a templated reply from support, and many people say a single email got them a refund. Refunding users is small change for it. What it fears is users going to their bank for a chargeback, because a high chargeback rate gets the Stripe account banned and the whole quiz page shuts down. So refund handling is part of the business too.

Where users come from

Only one thing can be confirmed: paid Meta and Instagram ads. Korean users say they clicked through an Instagram ad to a Korean page, and some say "the ad popped up and I just took an IQ test on a whim". How many ads ran and how much was spent can't be found from outside.

Almost no other channels show up. Searching cognius on TikTok turns up no videos, nobody discusses it on X, and on Reddit it is only named in a warning thread in r/iqtest. Instagram has a few scattered people posting their IQ scores (one Indonesian user posted 119), a very small amount.

Why did it surface in a dozen-plus countries at once? Because it translated the quiz page into 11 languages and ran ads by country. The app is just the step after payment, which is why reviews come in Korean, English, German, Russian and Spanish.

Who it's competing with

Cognius didn't invent this business. It is one of a large group of similar players:

What Cognius adds beyond a plain IQ page is 40 psychology tests, AI-written card-style readings, and an AI guide that remembers you. That makes it look more like a real product, and gives people a reason to "download the App." But almost nobody in the reviews mentions these features. They're all talking about the charges.

One more detail: the same Cyprus company has already applied this playbook to dog training with a Woof Method, "two-minute quiz, get a custom training plan." It looks like a team that specializes in quiz funnels, running one category and then moving to the next.

If it were you

C: you can learn something, but copying it isn't worth it. This is a mature gray-market play. Big players like myIQ and BetterMe have run it for years, and Cognius itself is just following the trend. Whether it makes money comes down to two things: whether you can run Meta ads, and whether you're willing to take on the risks of chargebacks, delistings, and account bans. The App itself has nothing worth copying.

Who it suits: small teams that already run Meta ads and can work out their acquisition costs and payback periods; or people with a subscription product who want to move part of their revenue off the App Store and onto the web.

Who it doesn't suit: indie devs who've never run ads and want to rely on organic App Store traffic; anyone hoping to win on product experience (users in this category pay because the ads push the test in front of them, not because the readings are good); anyone who can't stomach a screen full of 1-star ratings and refund emails.

What's worth learning is the structure, not the pricing:

1. web quiz → web payment → download the App to see results. Apple doesn't get its 30% cut, and on the web you can run A/B tests and multilingual landing pages freely. Off-the-shelf quiz funnel builders now exist, so you don't have to write your own.
2. "Pay a little to see your results" filters for people willing to pay better than a free trial does, and it captures payment cards along the way.
3. Use AI to turn a quiz result into a long, personalized reading, so people feel the money was worth it. The marginal cost is close to zero.
4. Copy the same quiz pages across categories: psychology, dog training, and the model also works for weight loss, language learning, and parenting.

A different approach: state the price clearly. The first screen says "Try for 3 days, then X per month, cancel anytime with one click." Set the monthly fee at a level users don't feel scammed by, and seriously work on retention inside the App. Conversion will be lower than Cognius's, but you won't be fighting chargebacks every day, and EU and South Korean consumer protection rules won't come after you. Those dozens of South Korean one-star reviews show that users in these markets will publicly complain, and regulators are strict about unclear auto-renewals.

How to acquire users: paid ads are the only route, mainly Meta and Instagram. The ad creative is the quiz question itself: "What's your IQ?" "What's your attachment style?" Start with a small-language market, where ads are cheap and there's less competition.

When to move: this play isn't tied to a time window, since similar operations have run for years. The real time limit is on Cognius itself. If its Stripe account gets banned, or Apple pulls the App over a flood of billing complaints, that could happen within months.

What would show we got it wrong: if six months from now Cognius's App itself keeps users, the US rating climbs into the thousands, and positive reviews start praising the AI guide and the quality of the readings, that would suggest "AI psychology self-tests plus companionship" has real demand beyond the ad funnel. At that point, a transparent-pricing, better-experience version would be worth a closer look, maybe even at a B or an A.