"Say I didn't get the ball? The heatmap doesn't lie": a Turkish app that fits pickup pitches with AI cameras climbed to #7 on the sports Top Grossing chart by signing up pitches in person

Akıllı Saha ("smart pitch") puts 4K cameras on small artificial-turf pitches in Turkey. AI handles the directing and produces movement data automatically, then turns every player into an FIFA-style player card. It launched in Sep 26 2026, has only 50 ratings on iOS, yet by Nov 6 it ranked #7 on Turkey's App Store sports Top Grossing chart. The idea isn't new: someone in Turkey ran a project with the same name and concept 2015 years ago. What lets it grow is a dozen-plus people signing pitches in person, plus a chart with a low bar. For anyone looking to follow it, this piece mainly explains what makes the business hard.

A video with 62k views

On Nov 2, @akillisaha.ai posted a TikTok video titled "Sahada Messi, raporda duran tabela!" ("Thinks he's Messi on the pitch, the report shows a signpost that never moves"). The caption goes on: after a pickup game someone always says "they didn't pass to me," and Akıllı Saha's data now "officially exposes" that, "the heatmap doesn't lie," and it ends by tagging the friend who always says it.

The account has about seventy followers, and most of its videos get around a thousand views. This one reached 62,078 views and 777 likes by Nov 6 (TikTok, 11-06). With so few followers and such high views on a single post, it was likely paid to promote (inference). Its next post on Nov 5 was titled simply "Halı sahadaki o arkadaşına gönder!" ("Send it to that friend of yours on the pickup team!") and got 7,015 views.

Both videos sell the same thing: the match gets recorded, the data gets calculated, then you use it to show up your teammates.

What it is

In Turkey, small artificial-turf pitches rented by the hour for 5-a-side to 7-a-side games are called Halı Saha, and booking one after work is a common habit. Akıllı Saha's product actually lives on the pitch: operator Paulabs sends people to partner pitches to survey, pick camera positions, install cameras, and calibrate. After that, each game runs like this:

• Scan the QR code at the pitch, or enter a match code a teammate shares, to join that game;
• Cameras record the full 4K pitch, and AI follows the ball and switches shots automatically ("auto-directing");
• After the game, scores, distance covered, speed, heatmaps and "player of the match" are produced, and the data is saved to the player's FC player card;
• Cards "level up" with play, and teams can enter online leagues at 6 tiers, one season every 3 months.

Pricing is per session (official pricing page, 11-06): a single-session analysis is 400 lira; packs of 4, 8 and 12 sessions work out to 350, 325 and 300 lira per session, and 12 sessions cost 3,600 lira total. The website takes payment through iyzico, and the app sells it too. A four-star review from Oct 3 says making player cards needs a separate premium: "nobody on the team agreed, the pitch fee's already paid, a subscription on top doesn't make sense." So besides the per-session analysis packs, there's probably a subscription layer too (inference).

The "About us" page on the company website describes a team of more than 15 people across 4 departments, two of which do offline work: "Technology & Installation" and "Pitch Operations."

Why it can climb the chart

First, signing one pitch brings a dozen-plus people a night who must install the app. To see footage of their own game, they have to go through the app. One pitch hosts several games a night, with ten to fourteen people per game, so customer acquisition rests entirely on installs and in-person signups. An Oct 19 five-star review says: "Heard about this from a friend today, I've wanted something like it for years." The negative reviews point to the same thing: "most cities don't have it" (10-29), "Arnavutköy in Istanbul doesn't have it" (10-11), "urgently need pitches at Sarıyer sports club" (10-22), "get to İzmir fast" (10-03). Users are queuing for it to sign up pitches near them.

The App Store numbers are small. The iOS rating 10-28 is 47, and 11-06 is 49. What's really growing is Android: Google Play installs 10-29 are 5,764,11-05 and 9,230, up 3,466 in a week (Google Play, daily numbers). Turkey has a lot of Android users, which fits the timing of the signing season.

Second, the bar on Turkey's sports Top Grossing chart is low. Around Nov 6, the apps near it were: #5 SkorAltyapı, #6 TipsPro AI, #8 Swimrank, #10 Goal Signal, and #11 a fishing weather app. From Oct 28 it has stayed between #7 and #17. One analysis costs 300 to 400 lira, so ticket sizes are decent, and a few hundred people paying per session is enough to reach that spot (inference). Ranking #7 only shows that people are paying inside the app. It doesn't reveal how much revenue there is, since Apple doesn't publish revenue.

Third, it packaged match footage as a game. Most competitors only offer recordings. It adds a card, a rank and a season, so players have something to show off and a reason to keep paying the following week. The social content almost writes itself: "who slept in the back line, who lay down up front, now exposed in seconds" (TikTok, 10-26). An Oct 14 review is interesting because it's from a wife: "I've become the VAR room at home. Mostly I use it to see how far my husband ran and the funny moments in the game... it made me want to set up a game myself."

It also paid for local content: the official Instagram has about 6,700 followers with posts tagged #reklam (ad); a local news account in Fethiye posted "FIFA player cards become real" on Oct 3; a Reel with a North Cypriot influencer @urban_cypriot got about 52k likes. On Oct 30 it entered Cyprus's sports free chart at #27 and by 11-06 had dropped to #99.

Who it's competing with

This has already been done in Turkey for ten years.

Overseas there are Veo, Pixellot and Trace, which sell hardware and annual fees to clubs and coaches, and don't touch hourly-rented pickup pitches.

Mackolik is the biggest variable. Its teaser came 10 days after Akıllı Saha launched, and the product is nearly identical: mid-line AI tracking cameras, auto-directing, and games posted to individual and team profiles. It has tens of millions of scores users, and it's going after the same pitches. Akıllı Saha's moat is only one thing: how many pitches it has already signed, and it hasn't disclosed that number.

What users are complaining about

Most of the October negative reviews on the Turkey App Store are about accuracy and stability:

• "Stats are wrong, scores don't show, every game gets marked as a draw" (10-11)
• "It has to recognize goals. If it can't, it's basically GG. The top speed is wrong too, it says my teammate hit 30 km/h, impossible. Total distance is also 500 to 600 meters short" (10-11)
• "After the update, no game loads at all" (10-14); "Can't even open the games, how are we supposed to pay for games we can't open?" (10-17)
• "Verification codes never arrive, can't even log in" (10-13, 10-21)

These problems show the AI recognition isn't stable yet, and inaccurate data goes straight to the core selling point of "settling it with data." On the other side, a later entrant could differentiate here by getting accuracy right.

If it were you

C. You can learn from how it works, but I don't recommend going head-to-head with it in Turkey. Copying it in another country is heavy lifting too.

• One-line verdict: This is a business that runs on hardware and in-person sales, and the app is just the payment gate. Turkey already has five or six players, and the biggest scores platform is coming in, so the window is closing. What's worth taking away is its cold-start structure: install equipment at offline venues first, then make people use the app to get their own content.
• Who it suits: teams that already have pitch, badminton court, pickleball court or go-kart track relationships in a city; those who can find cheap, stable AI tracking hardware (or partner directly with camera vendors) and have people to run offline installs and after-sales; and countries where people rent pitches after work for pickup games with no local players yet (check the Middle East, the Balkans, Latin America and Southeast Asia first).
• Who it doesn't suit: solo developers who can only write code and don't want to run offline operations. Without cameras installed on pitches there are no users and no content. It's also not for anyone hoping to take that market in Turkey, since there are too many competitors and you won't get Mackolik's traffic.
• How to build it: don't build hardware first. Find pitches and vendors that already have recording or streaming gear (Turkey's VAR Sistemi, which sells live streaming to pitches, is one example; other countries have similar companies), do only the software layer, meaning auto-cut highlights, stats, player cards and leagues, sell per session to players, and split revenue with the pitch. Accuracy is the make-or-break point: get goal detection, scores and distance covered right first, then work on the packaging.
• How to acquire users: sign pitches one at a time and put QR-code signs on each one; have teammates invite each other with match codes; shoot short videos that play up "roasting each other" (heatmaps, the player who ran least, player of the match), run small paid ad budgets, and write captions that prompt people to tag friends. The more pitches you sign, the better the chart position looks, and the easier the next signing becomes.
• When to move: if you're entering a new market, lock down the top pitches in one city before a local scores app or major sports platform launches the same thing. In Turkey, the landscape will be clear a few months before Mackolik Mahalle's official launch.
• What would prove this read wrong: ① After Mackolik Mahalle launched, Akıllı Saha fell out of the top 30 on the Top Grossing chart and weekly Google Play installs dropped below 1,000. That would mean venue sign-ups can't beat a traffic platform, and a small team can't win this way. ② The reverse: if it still holds the top 20 half a year after Mackolik launched, and its official partner venues reach a dozen-plus cities beyond Istanbul, then the people who signed venues first can hold their ground, and this grade should go up. ③ If its bad reviews keep stalling on "inaccurate data," then AI recognition is harder than it looks, and copying just the software layer will need more time and money.

Data from App Store Turkey and Cyprus charts and reviews (10-28 to 11-06), Google Play install counts (10-29 to 11-05), the akillisaha.ai site (11-06), TikTok @akillisaha.ai (11-06), public Instagram posts, Webrazzi (2016-04-12), and Mackolik's teaser on X (10-06). Anywhere marked "inferred" is an inference.