One-week trial for $1.99, then $29.99 a month: how chatbot.app climbed from 140k to 45k globally

A Turkish app studio bought the domain "chatbot.app", put GPT, Claude and Gemini into one chat window, and used ads to send people to a web checkout page. The product itself has nothing new, yet since September its global ranking has tripled. Below: who it belongs to, how it collects money, how it buys traffic, and what you should take from it and what to avoid.

Someone who thought they were paying in pesos

In Oct 25, an Argentine user gave "Chatbot App - AI Assistant" one star on Google Play: "I want to cancel my subscription. This app doesn't meet my requirements and I don't know how to cancel. I thought the price was in pesos, not dollars."

The same week, the review section from the same country had other comments: "It says I've used up today's quota, but I haven't used it at all, I just opened it", "It said there were free days, but as soon as I activated it, I was charged", "I can only send three or four messages a day." There were also five-star reviews: "Very useful, the answers feel more human than other AIs."

The app has 10.4M installs on Google Play, with about 7k new installs in the last 340 days, roughly 50k a day (Google Play, 11-07). It ranks #69 on the overall Top Grossing chart in Argentina, #78 in the Korea store, #82 in Brazil, and #85 in Mexico. It has charted in 28 countries in the productivity Top Grossing category, ranking #7 in Korea. The website chatbot.app, which shares the app's name, is the subject of this piece.

Who made it

The website's refund policy names the seller as "TapSuite Yazılım Hizmetleri A.S.", a Turkish company. TapSuite is a studio under HubX, a Turkish app group: HubX's careers page lists "Product Lead - TapSuite (İzmir)". In a web snapshot from July, the system prompt built into the chat window still read "You are Superbot, developed by the HubX team". HubX's Android apps together have about 71M installs, and TapSuite has its own regulars on the charts such as Scanner App and CleanX.

There's also a point of confusion: another nearly identical site, chatbotapp.ai, exists online with the same brand name, tagline and model list, and its terms say it is "an independent product of HubX Yazılım Hizmetleri A.Ş." So the same group is running two sites under two legal entities and two domains in parallel. Someone on Reddit posted "I got scammed by ChatbotApp.ai", but the link they shared was chat.chatbot.app, and users themselves can't tell the two sites apart.

The domain's history also shows it isn't a new product: in 2023 chatbot.app was still a parked page, and content appeared only around September 2024, which matches the launch dates of the Android version (2024-11-15) and the iOS version (2024-12-04). My inference is that HubX bought this category keyword domain in the second half of 2024; the purchase price wasn't found.

What drove the rise

Chrome's real-user statistics are even more direct: in May it was in the global top 100k, dropped from 6 to August to the top 500k, then jumped to the top 50k in September. By country, Vietnam entered the top 10k; France, the Netherlands, Indonesia, India, Brazil, Canada, the Philippines, the US, Mexico and Australia were all in the top 50k.

This is a push after it had already recovered, not a start from zero. Rankings are a 30-day rolling average and lag by a week or two, so the real volume surge began around mid-September. Over the same period, the sister site chatbotapp.ai stayed steady at around 40-50k globally, suggesting the group didn't shift traffic from the old site. Instead, both sites were funded together, and the new budget went mainly to chatbot.app.

How it collects money: the same plan at two prices

Open chatbot.app/pricing: Pro is $19.99 a month, Max is $39.99 a month.

On the ad landing page chatbot.app/landing/pricing, the same Pro reads "1-week trial for $1.99, then $29.99 a month", and Max becomes "from $69.99, coming soon" (11-07 web).

In other words, people arriving from ads see a different price set: first they pay $1.99 to lower the threshold for the first payment, then after a week it automatically converts to a monthly fee 50% higher than the official site's list price. Web checkout bypasses the app stores, so there's no 15–30% commission. The savings and the higher monthly fee let it afford more expensive traffic (inference).

The barrier for free users is kept very low: the homepage is the chat window itself, without login you can send only 1 messages, after which it's a login and paywall. The refund rule is "fewer than 20 messages sent, within 3 days of purchase." The review complaints about "quota used up right after opening" and "charged as soon as activated" are the other side of this funnel. Based on an estimate from the Peekly site's rating distribution, one-star reviews make up about 24%; meanwhile, each update note ends with "if you like it, please leave us a review", and many recent five-star reviews say only "excelente".

The third-party estimator Peekly says it earns roughly $500k a month with about 1M downloads a month, but the same page elsewhere says $116k over the last 30 days. The two figures contradict each other, so treat them only as an order of magnitude.

Where users come from

Almost all of it is bought, plus one good domain:

• Search ads: the site has Google ad click tracking and dedicated ad landing pages; the homepage also has a Microsoft Bing conversion pixel.
• Meta ads: Peekly shows about 100 ads running (unverified).
• New model landing pages: a new page opens for each new model: /gpt5.4, /claude-sonnet-5, /deepseekV4, /nano-banana-pro, /grok4.20 … the sitemap has 40 pages, capturing searches for "new model name + where to use it".
• Store search: the app title is "Chatbot App - AI Assistant" with package name ai.app.ai, claiming the words chatbot and AI assistant.
• The domain itself: chatbot.app is the category keyword, so people who type the URL directly, search "chatbot", or can't tell which is official might land there (inference).

Content channels are effectively absent: the TikTok account @chatbotapp has only 13 followers; nobody has mentioned it on X in the last 90 days; no media outlet has ever covered it.

Its main markets are places like Vietnam, Indonesia, the Philippines, India, Brazil, Mexico and Argentina. ChatGPT Plus is relatively expensive in these countries, local payments are inconvenient, ads are cheap, and users can't easily tell whether this is an official product.

Who it's competing with

This category is already saturated; in Turkey alone several studios run the same playbook:

On the web there are also multi-model aggregators like Poe (about 13k users globally) and Monica (about 15k users), which have their own brands and ecosystems. Above them sits ChatGPT itself. From 2025 to 2026, new wrappers such as Ocean AI, Genora, Zenda and TeraBot kept pouring ads into Meta.

Chatbot App isn't the biggest in this space. What sets it apart is three things: a category keyword domain; a web checkout route alongside the app, using a low-price trial to convert into a high monthly fee; and the same codebase sitting under several companies and brands, all running ads together. All three are about distribution and collecting money, not the product itself.

If it were you

Verdict: don't build the Nth multi-model chat wrapper. The business makes money, but it runs on ad budget, a purchased good domain, and a whole group's media-buying experience, none of which followers have. A dozen-plus players are already spending in the category, and the price war has long started. What's worth taking is its way of collecting money: separate pricing on the ad landing page, and web checkout.

Who it's for: Teams that already have an app earning from subscriptions and paid acquisition (tools, photo editing, learning apps all fit), with a lot of Android users, and that are paying store commission.

Who it's not for: Solo indie devs with no paid acquisition experience and no launch budget; people hoping product differences will win (users in this category can't tell products apart); anyone uncomfortable with charging the card right after activation. Part of this funnel's conversion comes from exactly that.

How to do it:

1. Build a web version of your app and a separate ad landing page. Test a "one-week trial at a low price, then monthly subscription" offer on the landing page, priced separately from the store listing.
2. Take web payments through Stripe or local payment methods, and test in one or two emerging markets (Vietnam, Mexico, Brazil). Ads are cheap there, and you can measure directly how willing users are to pay on the web.
3. If you're building an AI tool, launch a landing page the same day each new model drops, and target search for timeliness.
4. Spell out your refund policy and show prices in local currency. That commenter who thought it was pesos: on Stripe, that becomes a chargeback.

How to get users: Google / Bing search ads paired with dedicated landing pages; Meta ads targeting Android users in emerging markets; new-model landing pages for long-tail search. Don't count on TikTok or communities. They're close to zero for this product too.

When to move before others: Bypassing the store with web checkout depends on Apple and Google's policies on external payment links, which are currently loosening. If you get the web funnel running in the first half of 2027, you can capture the commission gap while it lasts.

What would show you're wrong: If by the end of 2026 chatbot.app's global ranking drops back below 100k and Android daily installs fall under 10k, the extra spend didn't pay off. Refunds and chargebacks from the low-trial-to-high-monthly switch in these markets ate the profit. At that point, don't copy this funnel as-is. At minimum, narrow the gap between the trial price and the monthly price.