
A Turkish-language Cal AI clone: EatBetter used fitness creators' ad videos to take monthly revenue from $10k to $180k in two months
What it is A subscription app that uses AI to calculate calories and nutrition from a photo of your plate. It is Turkish-first and knows local dishes. Its website is a food calorie site in 5 languages that funnels traffic to the app.
Why it took off It paid Turkish fitness creators to promote it with discount codes, then ran their videos that worked as Meta ads while changing the paywall. Monthly revenue went from $10k to $180k in two months. Once the numbers added up, it raised $2M in paid-acquisition funding without giving up equity. The website's recent climb came from the multilingual calorie pages launched in July. Their effect on app growth isn't visible yet.
BHow you could follow it Pick a mid-sized language market that the big global players haven't localized seriously and where Cal AI doesn't rank in the top 30. Build a local-dish version in the Cal AI mold, test it with local creator discount codes, then take the videos that bring in paying users and scale them through Meta testing. Build the calorie website last.
Global rank 264k · 90 days ago, top 1M+
A small team in Izmir built a photo-based calorie app that has sold almost only in Turkey and reached #1 in the Health & Fitness Top Grossing chart on the Turkish App Store. In July this year its website became a 5-language food calorie database, and its global ranking tripled within a month. Below: how the money was made, whether the website's climb actually helped, and whether this playbook could work again in another country.
A fitness video labeled "ad"
In Feb 2025 21, Turkish fitness creator @fit.kultur (130k TikTok followers) posted a video whose caption opened with "REKLAM VE TAVSİYE": advertisement and recommendation.
One of the biggest problems with calorie-tracking apps is that when you search a food, dozens of options pop up and you can't tell which is right. Eat Better aims at exactly that: snap a photo of your meal, add a note on the portion, and AI works out the calories and nutrition.
The last line was a discount code: join with FITKULTUR for 7% off. The video got 317k views. More than a year later, in April 2026 727 he posted another, with 11-11k views (checked on TikTok).
There was more than one of these videos. The personal trainer @pt.samedcansengonul used the code SAMED30, @sudeozkkull used SUDEOZKUL30, and @eraysirinn used ERAY30. Each creator had a "name + 30" code, all tagged #reklam. In October 2025 the website even published a revenue-share agreement page. The official TikTok account had only 222 followers, and other people did all the talking.
What happened next is described by Turkish growth consultant Samet Durgun on his own case study page (growththerapist.com, 2026-09-07): in 2025 he took over EatBetter's Meta ad account. He found the creator videos "already working in organic reach, with paid not yet caught up", so he tested them one by one as ad creative, cut the ones that didn't run, and spent more on the winners. He also changed the paywall. Over two months, monthly subscription revenue rose from $10k to $180k, and the ad payback rate went from 70% to 230%.
What it is, and where the money comes from
The app is called EatBetter: AI Food Diary. Photograph a plate and the AI identifies the food, estimates the portion, and gives calories and the three macronutrients. There is also an AI nutrition coach named Betty you can chat with, plus water, step, and fasting tracking and a community where you can look at other people's plates. Download is free. Pro membership unlocks unlimited photo scans and chat. It launched on iOS in 2024 September. The seller is the US shell company AI Sparks Labs, LLC, and behind it is BETTER LABS in Izmir. Founder Murat Yaşar previously did iOS and product work at Turkish secondhand marketplace Gardrops.
It's almost exclusively a Turkey app. App Store ratings: Turkey 19,491, Germany 252, US 224, Azerbaijan 208, UK 77 (11-11). Revenue is only visible in Turkey too: 10-28 It ranks #1,11-11–#7 on the Turkey App Store Top Grossing chart in Health & Fitness, and occasionally hits the Turkey overall Top Grossing chart at #140–#180; no other store has it charted. On Google Play it has 199K+ installs in total, with a rating of 4.82.
Pricing shows up in user complaints: one person subscribed on 24 September and was charged 899.99 lira at once; another bought an annual plan for about 1,200 lira; there is also a three-month tier (Turkish complaints site şikayetvar, as of November).
In 2026 July, Turkish paid-acquisition funder Leus Capital gave it $2M. No equity was taken. The money is advanced against forecast returns from each cohort of users (mobidictum, 2026-07-14). The founder's words: "Our growth math has worked for a long time. Every cohort we buy pays back. The bottleneck was never the product or the channel. It's how fast the money turns over."
How it took off: three steps, each one visible
- Let creators talk first. Turkey has fitness and weight-loss creators at every size, from 130k followers to personal trainers with a few thousand. They were paid per post, each with their own discount code. You can see which creator drove how many sales, which gives you a number to judge whether a video worked.
- Then run the videos that work as Meta ads. The videos look like the creators' own posts, not ads. The consultant tested them like regular ad creative and only added budget to the winners. The paywall was changed at the same time. The case page says a small lift in conversion mattered as much for payback as swapping in a better ad.
- Once the numbers add up, borrow to scale. If each cohort pays back, you can borrow paid-acquisition money against users' future revenue instead of selling shares.
Why could it reach the top in Turkey? The biggest global players haven't seriously targeted it. Cal AI ranks only #34 in Turkey's Health & Fitness grossing chart, and MyFitnessPal is #73 (11-11). A product that is Turkish-first, knows simit, börek, lahmacun and ayran, and speaks through Turkish creators squeezed in.
But don't read this as "nobody in Turkey is doing it". On the same Turkish Health & Fitness grossing chart, 5 of the top 10 are calorie apps. FoodPilot, at #1, is a Hong Kong company that grew purely through paid acquisition. Yazio is #2, Foodvisor #4, and the newcomer Welmi #10 over the same period. Localization got it into the front rank, but it didn't make it dominant.
Why the website suddenly climbed: one lahmacun page
What drew attention to it this time was actually the website. eatbetter.app was just an app landing page for the first two years. In 2026 July 17 the homepage went from 4KB to 15KB, and that same month the web archive first captured thousands of food pages (Wayback). It is now a food calorie database in five languages: Turkish, English, German, Spanish and French. The homepage says "347,427+ foods" and cites the USDA as its data source. There are also comparison pages, lists such as "high-protein foods", and calorie, BMI and protein calculators.
Open the English lahmacun page and the title reads "Lahmacun Calories: 396 in 1 medium (180 g)". Below is a nutrition table, and below that a set of Q&As: "Is lahmacun the same as Turkish pizza?", "What's the difference between lahmacun and pide?", "How do you pronounce lahmacun?" The answer is lah-mah-JOON, "the Turkish c is pronounced like the j in jam". The page carries App Store and Google Play buttons, and every link carries a tag identifying its source page. A sticky bar across the site reads "Start losing weight · Create My Plan", and clicking it opens a weight-loss plan questionnaire that starts by asking your age.
Global rank changes (Tranco): 10-13 from #750k → 10-27 at #410k → 11-10 at #270k; 90 days ago it was outside the top 1M. In Chrome real-user rankings, Turkey rose from the top 500k in July to the top 50k in October; Germany, the UK, France, Spain, Australia and Canada, plus Indonesia, Vietnam, South Korea and Mexico, all entered the top 500k between 9 and October.
What kind of search is it eating? A Brave spot-check: in the German store, searching "ayran calories" ranks it #3; in the UK store, "doner kebab calories" ranks it #6 and #8; in the US store, "lahmacun calories" ranks it #8. But for big terms like "calories in egg" and "calorie calculator", it doesn't show in the top 20. In Turkish, "simit kalori" and "yumurta kaç kalori" are held by local incumbents Diyetkolik and FatSecret (Brave is not Google, so treat this as directional only). In other words, it owns the long tail of "looking up Turkish food calories in a foreign language". Many of the people searching ayran calories in Germany or the UK are Turkish.
Two small things show this site was built by people who know the trade: robots.txt blocks the crawlers of SEO tools like Ahrefs, SE Ranking and DataForSEO, so rivals can't see its backlinks and keywords. Comparison pages, which can be combined endlessly, are blocked by default, and only a few dozen valuable ones are allowed. ChatGPT, Perplexity and Claude's search crawlers are allowed, while training crawlers are blocked. It wants to be cited by AI search, but it isn't giving away its training data for free.
But this recent climb in website traffic doesn't show any sign of bringing in app users. Over the same period, Google Play gained only 100–250 installs a day (10-29 to 11-11 went from 196,707 to 199,261), and the Turkish App Store added just 91 ratings in a week. Maybe the traffic hasn't been monetized yet, or conversions are happening on the web through the questionnaire and payment (robots.txt blocks /payment, and complaint sites mention EatBetter charges showing up on carrier bills, which suggests web checkout). From the outside you can't tell. The safe reading: this is a paid-acquisition app adding a cheap channel for itself, not the reason it took off.
What users are complaining about
On şikayetvar, EatBetter's satisfaction rating is 22/100. Complaints cluster around two things: emails getting no reply, and charges continuing after cancelling a subscription (roughly 35% of complaints). The gist of one complaint:
I bought three months in the App Store, cancelled in the app in August, confirmed the cancellation, then deleted the app. Without my consent it renewed me through Nov 30.
Subscribed in Sep 24, 899.99 lira charged to my card. After a day I found that apart from the photo feature, there's basically nothing in it.
This is the common problem with paid-acquisition subscription apps, and it's an opening for followers: clean cancellations and refunds are a selling point in a market built on blogger word of mouth.
Who it's competing with
| Competitors | Scale | How it differs |
|---|---|---|
| Cal AI | Play 4.58M installs; Turkey health top-grossing #34 | Category benchmark, mainly English and US |
| FoodPilot (a Hong Kong company) | Turkey health top-grossing #1 | Pure paid acquisition, no local bloggers or local food |
| Yazio | Play 58.9M installs; Turkey iOS 61k ratings, health top-grossing #2 | Long-established European player, already ran multilingual calorie-table sites |
| Foodvisor | Turkey health top-grossing #4 | Long-established French photo recognition |
| Welmi | Play 1.86M installs, +39k in a week; Turkey health top-grossing #10 | A 2025-year-old newcomer with more aggressive paid acquisition |
| FatSecret, Diyetkolik | Consistently top few for Turkish calorie-term searches | An old-school lookup site, far stronger in search than EatBetter, with no AI photo recognition |
| Nutrola | Claims 2M users on its site; global rank went from 90k to 860k in 330 days | Also an "AI calorie app + multilingual content site", the closest match in approach, and also has a Turkish-language blog |
| Kilo: AI Calorie Counter | Turkey iOS 500 ratings | Founded in 2025 November, same concept built by a Turkish indie developer |
| Photo recognition via ChatGPT / Gemini | — | Free, but no logging, no goal setting, no reminders |
If it were you
B. Worth a try: the "local version of Cal AI for one country" path works, and EatBetter lays out every step. But it's a paid-acquisition business, and the category is one of the most crowded subscription niches of the past two years. Localization won't stop outsiders who are willing to spend on ads.
- Verdict: what's actually valuable is its playbook: local bloggers test with codes → turn the videos that work into Meta creatives → test paywall changes at the same time → borrow money to buy more acquisition once it pays back. Followers can copy all of this without big-company traffic or exclusive resources. Its SEO site is icing on the cake, and its contribution to revenue hasn't been proven. Don't copy it as the main course.
- Who it suits: a team of 1–3 people, at least one of whom has run Meta ads and can work out the acquisition cost of each subscription. Or someone who speaks a mid-sized language market's language natively or lives there, where global big players have not localized seriously but people will pay for apps, and who can personally recruit local fitness bloggers.
- Who it doesn't suit: anyone who doesn't run ads and is counting on app store or SEO organic traffic; anyone who wants to head-to-head Cal AI and Yazio in English or German markets; anyone who wants to build another one in Turkey directly (5 calorie apps are already in the top 10 health top-grossing).
- How to do it: pick a market by checking the charts first. How many calorie apps are in the top 20 of the local App Store health & fitness top-grossing chart, and where does Cal AI rank? Enter only where Cal AI is outside the 30 spot, local cuisines are complex, and local-language calorie-table sites are still very old (Vietnam, Indonesia, Poland and Romania are worth a look, though this article hasn't verified each one). Build the product like Cal AI: photo recognition, with portion sizes and calories for local dishes calibrated separately (the selling point EatBetter's blogger videos push is that searching one food pulls up dozens of options). Offer three subscription tiers: weekly, three months and yearly, with cancellation and refunds spelled out clearly.
- How to get users: first find 10–20 local fitness and weight-loss bloggers, anywhere from a few thousand to a few hundred thousand followers, each with a "name + 30" discount code. Track which videos actually bring paying users, then take those few to Meta as ads and test them for a few hundred dollars per round, scaling up only the winners. Once it pays back, you can look for revenue-based acquisition financing (companies like Leus) rather than selling equity. Once the app is established, learn from EatBetter's "local food × foreign language" calorie pages to catch search traffic from overseas expats.
- When to start: weight-loss apps peak in January each year. Launch and start recruiting bloggers by mid-December, and lock in creatives that pay back during January. If you miss it, the next window is 4–May, before summer.
- Signs you've read it wrong: ① after EatBetter took in $2M, by February 2027 it actually dropped out of the top 10 of Turkey's health top-grossing chart, which suggests more money doesn't buy enough demand and the ceiling of this path is low; ② outsiders like FoodPilot and Cal AI start mass-recruiting Turkish bloggers and running ads with Turkish creatives, pushing EatBetter's ranking down, which means localization is not a moat, just an early move; ③ the opposite case: EatBetter's Google Play daily installs clearly rose in early 2027 and its website ranking kept climbing, which would mean the multilingual calorie site really does bring users, and it should move from "icing on the cake" to step one.
Data and sources
- Blogger videos and discount codes: public videos and captions from TikTok @fit.kultur, @pt.samedcansengonul, @sudeozkkull, @eraysirinn, @lifeofulya and @eatbetter.app (11-11 checked).
- Revenue and ad spend: Samet Durgun's case study page growththerapist.com/case-studies/eatbetter (2026-09-07; a consultant's self-reported figures for two months in 2025; later revenue is not public).
- Funding: mobidictum (2026-07-14), Leus Capital and the founder's LinkedIn posts.
- Ratings and charts: App Store ratings counts by country and the Turkey health & fitness top-grossing chart (10-28 to 11-11); Google Play installs and ratings (10-29 to 11-11).
- Website: eatbetter.app homepage, lahmacun food page, robots.txt (read 11-11); Wayback web archive; global ranking Tranco (10-13 to 11-10); Chrome real-user traffic bands CrUX (2026-05 to October).
- Search rankings: Brave Search spot-check, not Google rankings.
- Complaints: the EatBetter page on şikayetvar (as of November).
- Where the text says "inferred", "estimated" or "unverified", those are inferences, estimates and unverified claims.
Research notes: competitor table, acquisition evidence, full research report
- Competitors
Name Where Launched Scale How it differs from this one Cal AI iOS 2024-04 Play 4.58M installs; calai.app Tranco 147,987; TR health top-grossing #34 Category benchmark, mainly English and US; EatBetter is the Turkish-first local version AI Calorie Tracker – FoodPilot iOS 2024-11 TR health & fitness top-grossing #1 Hong Kong company, pure paid acquisition; EatBetter has local creators and a Turkish food database Yazio iOS 2014–2015 Play 58.9M installs; TR iOS 61,429 ratings, health top-grossing #2; yazio.com Tranco 33,061 Long-established European player with built-in calorie-table SEO; EatBetter is a size smaller, with the same SEO playbook Foodvisor iOS 2016 TR health top-grossing #4; foodvisor.io Tranco 127,001 Long-established French photo recognition, not targeted at Turkey Welmi Android 2025 Play 1.86M installs, +39k in 7 days; TR health top-grossing #10 A newcomer of the same period, with more aggressive paid acquisition and far more Android installs than EatBetter MyFitnessPal Other 2009 Play 137M installs; Tranco 4,459; TR top-grossing #73 A giant, mainly manual logging with Meal Scan; weak paid performance in Turkey FatSecret(fatsecret.com.tr) Web Around 2007 Play 63.2M installs; consistently in the top 5 for Turkish food calorie searches A veteran of Turkish calorie-table SEO; EatBetter's website runs head-on into it. It has no localized AI subscription playbook Diyetkolik Web 2012 Play 1.03M installs; diyetkolik.com ranks #2 for simit / yumurta calorie terms A Turkish local lookup site with stronger SEO but no AI photo recognition Nutrola Web 2025 nutrola.app Tranco 334,320 (90 days ago 861,680); the site claims 2M users The same kind of app + multilingual content site, aimed globally; EatBetter focuses on Turkish food ChatGPT / Gemini photo calorie estimates Big-company features 2024 General assistant Free, but no logging, goals or reminders - Where users come from
- Verified Turkish creators paid to promote it (TikTok / Instagram): @fit.kultur (130k followers) posted it labeled "REKLAM", 2025-02 317k views, and another 2026-04 727k views; a PT creator's discount code SAMED30; Instagram @eat.betterapp with 17k followers and lots of #reklam #ortaklık posts; the site has an affiliate-agreement page at 2025-10
- Verified Meta ads (creator footage) + paywall optimization: per the growththerapist.com case study, a consultant ran the Meta account for 2025 years. Creator videos were tested in a knockout format, then scaled. MRR went from $10K to $180K, blended ROAS from 70% to 230%. Our ad library does not track it.
- Verified UA financing to scale paid acquisition: Leus Capital 2026-07 provided $2M in non-dilutive UA financing (mobidictum 2026-07-14, Leus LinkedIn).
- Verified Programmatic SEO (multilingual food nutrition database): Wayback: 2026-07 homepage grew from 4KB to 15KB, and it first captured 957 /tr/besin pages and 511 /en/food pages. Brave spot-checks: ayran calories (DE) #3, doner kebab calories (GB) #6/8, lahmacun calories (US) #8. Broad head terms are not in the top 20. Tranco: ×3.5 since October.
- Inferred Web quiz funnel / web checkout: a site-wide floating bar, Create My Plan → /en/plan quiz. robots.txt blocks /payment. Sikayetvar has Paycell charge complaints.
Why EatBetter (eatbetter.app) took off
In one line: EatBetter is an AI photo calorie-tracking app built by a team in Izmir, Turkey (founder Murat Yaşar; company BETTER LABS YAZILIM, US shell AI Sparks Labs, LLC). The app itself launched in September 2024, but it really took off through Turkish local creator content → turned into Meta paid ads + paywall changes (2025; the consultant says MRR rose from $10k to $180k in two months). In July 2026 it raised $2M in paid-acquisition financing from Leus Capital. The website ranking climb that our recall surfaced is a separate story: from July 2026 the site was rebuilt as a programmatic nutrition database (SEO site) in 5 languages, claiming 347k foods. It has been climbing Tranco since October, and its long-tail traffic mostly comes from "Turkish food + English/German/French/Spanish" queries. Whether the SEO site drove any App downloads is not visible yet.
---
1. What it is and where its traffic lands
- App: snap a photo of your plate → AI identifies the food, estimates portions, and gives calories and macros. Also an AI nutrition chat called "Betty", water/steps tracking, intermittent fasting, and a community (follow other people's plates). Free to download; Pro subscription unlocks unlimited photo analysis, unlimited chat and full history (site FAQ, www.eatbetter.app/en). One Turkish-store annual subscription charge of 899.99 TL drew a user complaint (sikayetvar).
- iOS: listed as id6639614109,2024-09-17, seller AI Sparks Labs, LLC. Rating counts: TR 19,491, US 224, DE 252, AZ 208, GB 77 (iTunes lookup, 2026-11-11). Essentially a Turkish app.
- Play: com.aisparkslabs.eatbetter, listed 2024-09-26, 199,261 installs cumulative, +7 installs in the last 736 days, 100–250 per day (googleplay library app_day, 10-29 → 11-11), rating 4.82 / 11.3k.
- Where the revenue is: App Store Turkey Health & Fitness Top Grossing, #1–7 (10-28 #1,11-11, #7), occasionally #140–180 on the overall Turkey Top Grossing chart; Play Turkey Health Top Grossing #15 (appstore / googleplay libraries). No other country charts it.
- Website eatbetter.app: Next.js, 5 languages (tr/en/de/es/fr). Homepage says "347,427+ foods" and "Turkish & global cuisines", with data sourced to USDA FoodData Central. It has food pages (/en/food/…, /tr/besin/…), category pages, brand pages, restaurant pages, comparison pages (X vs Y), food lists (high protein foods…), calculators (calories/BMI/TDEE/protein/macros/body fat), meal plans and a blog. The whole site pushes to the app (store buttons with UTM tags) and a web quiz funnel at /en/plan ("Start Losing Weight / Create My Plan", starting with an age-range question). robots.txt blocks /payment, which suggests a web checkout (not verified).
- Traffic mainly lands on web content (SEO), while the money comes from Turkish iOS subscriptions. There is no public evidence of the conversion between the two.
2. Growth data
| Metric | Data | Source |
|---|---|---|
| Tranco | 10-13 750,236 → 10-20 586,078 → 10-27 413,075 → 11-03 304,529 → 11-10 269,511; not in the top 90M 1 days ago | traffic library, rank_day |
| CrUX Turkey | 2026-05–07 top 500k → 08–09 top 100k → before October top 50k | traffic library, crux_domain |
| CrUX other countries | 9–October new to top 500k: de gb in au ca ph (from September); us es fr it nl pl kr mx th vn id (October). br was top 1M before October; global top 500k before October. | Same as above |
| iOS TR rating | 19,290 (10-28) → 19,491 (11-11); 7 days +91, 30 days +201 | appstore library |
| Play installs | 196,707(10-29)→ 199,261(11-11) | googleplay library |
Read: the site's ranking roughly 3.5x'd, but the app side barely moved over the same period (Play gaining a couple hundred installs a day, TR iOS a dozen ratings a day). So this Tranco climb is organic traffic from a content site getting indexed, not an app surge. Among the countries newly ranking on CrUX are kr/th/vn/id/mx/in, which have nothing to do with Turkish diaspora, so the English generic food pages are pulling traffic too, not just Turkish dishes.
Timeline (Wayback CDX): in 2021 the domain was just a redirect page; from 2024-09 it was an app landing page (homepage about 4KB); in 2026-07-17 the homepage became 15KB, and the same month Wayback first captured large numbers of /tr/besin/ (957 pages) and /en/food/ (511 pages). /en/foods and /es/alimentos appeared 10–November. → The SEO site went live in July, and Tranco began reflecting it two or three months later, consistent with Tranco lag.
3. Competitors
| Product | Where | Launched | Scale | How it differs from EatBetter |
|---|---|---|---|---|
| Cal AI | iOS/Android | 2024-04(iOS) | Play 4.58M installs; calai.app Tranco 147,987; TR Health Top Grossing only #34 | Category benchmark, mainly English / US-focused, driven by US creators + TikTok. EatBetter is the Turkish local version. |
| AI Calorie Tracker – FoodPilot | iOS | 2024-11 | TR Health Top Grossing #1 | Hong Kong company, pure paid-acquisition model. EatBetter has local creators and a local food database. |
| Yazio | iOS/Android/web | 2014–2015 | Play 58.9M installs; TR iOS 61,429 ratings, Health Top Grossing #2; yazio.com Tranco 33,061 (already has kalorientabelle SEO pages) | Established European major, started in German. EatBetter is an order of magnitude smaller, and its SEO playbook is one Yazio has been running for a long time. |
| Foodvisor | iOS/Android | 2016 | TR health top-grossing #4; foodvisor.io Tranco 127,001 | Established French photo-recognition app. EatBetter is more local. |
| Welmi | iOS/Android | 2025-02(iOS) | Play 1.86M installs, +39k in 7 days; TR health top-grossing #10 | A rising newcomer at the same time, buying traffic more aggressively. Its Android numbers far exceed EatBetter's. |
| MyFitnessPal | iOS/Android/web | 2009 | Play 137M installs; myfitnesspal.com Tranco 4,459 | The giant, with Meal Scan. TR Top Grossing only #73. |
| FatSecret | All platforms | Around 2007 | Play 63.2M; fatsecret.com.tr ranks top 5 for Turkish food calorie searches | An veteran in Turkish-language food-calorie SEO. The EatBetter site runs head-on into it. |
| Diyetkolik | iOS/Android/web | 2012 | Play 1.03M; diyetkolik.com ranks #2 for "simit kalori" and "yumurta kaç kalori" | A long-established Turkish local calorie-table site. Much stronger than EatBetter in SEO, but with no AI photo recognition. |
| Nutrola | iOS/Android/web | 2025 | nutrola.app Tranco 861,680 → 572,468 → 334,320; the official site claims 2M users | The same "AI calorie app + multilingual content site" model, and it also has a Turkish-language blog (tr.nutrola.app). Its approach is the closest match. |
| Kilo: AI Calorie Counter | iOS | 2025-11 | TR 500 ratings, Health Top Grossing #69 | A same-style app by a Turkish solo developer. Small. |
| ChatGPT / Gemini photo calorie estimates | Big-company features | 2024– | — | Free and generic, but no logging and no plans. Subscription apps sell "logging + goals + reminders". |
Verdict: there is no shortage of competitors. AI photo calorie tracking is one of the most crowded subscription categories in 2024–2026. Within the top 10 of the Turkey Top Grossing chart, there are already 5 calorie apps.
4. How users came in (growth levers)
- Turkish fitness/weight-loss creator content, paid partnerships labeled "REKLAM" (confirmed)
- TikTok @fit.kultur (130k followers) posted 2025-02-21 "REKLAM VE TAVSİYE" promoting EatBetter, 317k views; 2026-04-16 posted another with 727k views. PT @pt.samedcansengonul shared a discount code "SAMED30" for 2026-05 30% off. Also mid-tier accounts like @lifeofulya, @handeozuss, @sudeozkkull (TikHub search 2026-11-11).
- Instagram account @eat.betterapp has 17k followers, with lots of "#reklam #ortaklık" partnership posts and user weight-loss stories (Jina search results). The website 2025-10 had /affiliate-agreement in the Wayback archive, which means there's a creator affiliate program.
- The official TikTok @eatbetter.app has only 222 followers. The main channel is creators, not the brand account.
- Creator videos run as Meta ads + paywall changes at the same time (confirmed, from the vendor)
- Case study page from growth consultant Samet Durgun (growththerapist.com/case-studies/eatbetter, published 2026-09-07): in 2025 he personally ran the Meta account, turning creator content that "already worked organically" into ads, testing creatives in a knockout format while optimizing the paywall. MRR went from $10k to $180k in two months, blended ROAS from 70% to 230%.
- It is not in our Meta / Google / TikTok ad libraries (not on the list; not appearing there doesn't mean it wasn't running).
- Paid-acquisition financing to scale (confirmed)
- 2026-07 Leus Capital provided $2M in non-dilutive UA financing (disbursed based on cohort payback forecasts), reported by mobidictum.com 2026-07-14 and hubogi.com. The founder's words: "Every cohort is paying back; the bottleneck is how fast the capital turns over." Leus announced it on LinkedIn, and the founder reshared it (19 likes).
- Programmatic SEO content site (confirmed the site exists and ranks; its contribution to the app is not verified)
- 2026-07 launched, in 5 languages × hundreds of thousands of food pages + comparison pages + checklists + calculators. Brave spot checks: "ayran calories" (German store) #3, "doner kebab calories" (UK store) #6 and #8, "lahmacun calories" (US store) #8, "lahmacun kaç kalori" (Turkish store) #9. But on big head terms like "simit kalori", "yumurta kaç kalori", "calories in egg", and "calorie calculator" it doesn't show up at all 20. It wins on the long tail of Turkish food in foreign languages; the generic big terms still belong to FatSecret, Diyetkolik, and Healthline.
- The robots.txt is deliberate: comparison pages are blocked from crawling by default, with only a few dozen high-value combinations allowed through the whitelist (to control crawl budget and avoid thin content). It blocks AhrefsBot, SERankingBacklinksBot, and DataForSeoBot (so competitors' SEO tools can't see its backlinks and keywords). It allows AI search crawlers like ChatGPT-User, PerplexityBot, and Claude-SearchBot, and blocks training crawlers like GPTBot and ClaudeBot. It wants to be cited by AI search without giving away its training data.
- Web quiz funnel (entry point seen, rest not verified): sitewide banner "Start losing weight · 4.8 · 18K ratings · Create My Plan" leading to the /en/plan quiz. This is the typical web 2app landing page setup used for Meta ads. sikayetvar has a user complaint, "Paycell account charged for EatBetter mobile payment," which suggests a web or carrier billing channel that bypasses Apple's cut.
5. Why it took off
- The Turkish market is underserved by global AI calorie apps: Cal AI ranks only #34 on the Turkish Top Grossing chart, and MyFitnessPal #73. A Turkish-first product that understands local food (simit, börek, lahmacun, ayran) and uses local creators could crack the top 7 in Health & Fitness. Evidence: TR iOS 19,491 rating vs 224 in the US.
- Creator content validated organically first, then moved wholesale into paid ads, combined with a paywall redesign that pushed ROAS past the line (case page 70% → 230%). Once that worked, UA financing scaled it: the classic "prove cohort payback first, then borrow to buy installs" playbook.
- The founder is a growth-focused product person: Murat Yaşar previously did iOS and product at Gardrops (a Turkish secondhand e-commerce app with millions of users). His LinkedIn says he took the app's rating from 34k to 100k+ in a year, and the team brands itself as "product-led growth".
- The recent website climb came from programmatic SEO launched in July, which took the underserved niche of "Turkish food calories in foreign languages" and added English generic food pages on top. Timing-wise, it's a low-cost acquisition channel added after the paid-acquisition engine was already funded.
6. Differentiation
- For global players: localization. Turkish-first product and support (though sikayetvar satisfaction is only 22/100), Turkish food data, Turkish creators, and local payments. This is the main reason it makes money in Turkey.
- For local Turkish calorie sites (Diyetkolik, FatSecret TR): it has AI photo logging and subscriptions, while they are old-style lookup databases.
- On the website side: a "Turkish food × multilingual" nutrition database is an SEO differentiator, but the programmatic food-database approach itself isn't new (Yazio, FatSecret, Nutrola, calory.app, and nutriscan.app all do it). Its contribution to app revenue hasn't been proven.
- Bottom line: differentiation (localization) is why it works in Turkey. The website climb we can see is an SEO channel added by an app that was already working on paid acquisition. It is not why the app took off.
7. Visibility and copycats
- Revenue has been disclosed: a growth consultancy's public case study reports MRR going from $10K to $180K (2025). Leus's $2M UA financing appears in Turkish mobile-industry outlets like mobidictum and hubogi. But the discussion footprint is narrow: in x-radar's corpus, "eatbetter" appears 0 times over 365 days, with no English tech press or major X accounts covering it. TrustMRR has no record.
- Copycats and peers: Nutrola also runs a multilingual content site and is climbing. Kilo: AI Calorie Counter (2025-11), by a Turkish solo developer, is a local lookalike. The bigger picture is a crowded AI calorie-app market (FoodPilot, Welmi, WiseMeal, and DietAI are all bigger than it on Play).
8. What others can learn
- Find mid-sized language markets that global apps ignore (like Turkey: large population, modest subscription spending, weak competition), then break into the top charts with local creators and local content.
- Creator content → Meta ads → paywall redesigned together. Get cohort payback working first, then scale with cohort-backed UA financing (Leus, PvX, Braavo-style) without diluting equity.
- Build SEO sites around the "local food × foreign language" long tail: you can't beat FatSecret or Healthline on head terms, but local-dish queries in foreign languages like "ayran calories" and "lahmacun calories" are almost unclaimed. The approach can be copied to Vietnamese, Indonesian, Persian, Brazilian, and other regional cuisines.
- Robots details: blocking SEO-tool crawlers hides the playbook; the whitelist controls combinatorial page explosion from comparison pages; allowing AI search crawlers.
9. Still unclear
- How much SEO traffic converts to app downloads or subscriptions: app-side data hasn't moved with it, which could mean the traffic isn't monetizing, or that conversion happens through web checkout (which we can't see).
- Whether there is a web paywall (/payment is blocked in robots, plus the Paycell complaint), and how large a share it is.
- Current MRR for 2026: the $180K figure is a consultant's account of two months in 2025. There's no data on whether it held up or how large post-funding paid acquisition got. Our ad libraries don't track it.
- The official site's numbers don't match each other ("2M+ daily meal analyses", "100,000+ people", "347,427 foods", and the same page also says 140,000+ food items). Credibility is moderate.
- A Brave ranking isn't a Google ranking; the SEO spot checks are directional evidence only.
Sources
- Official site: www.eatbetter.app/en, /en/plan, /robots.txt (read 2026-11-11)
- Wayback CDX:web.archive.org/cdx/search/cdx?url=eatbetter.app/*
- Growth consultancy case study: growththerapist.com/case-studies/eatbetter (2026-09-07)
- Funding: mobidictum.com/leus-capital-eatbetter-royo-games-ua-financin… (2026-07-14), www.hubogi.com/eatbetter-ve-royo-games-leus-capitaldan-34-mi…, Leus and founder LinkedIn post www.linkedin.com/posts/muratyasarr_proud-to-have-leus-with-u…
- Company: www.linkedin.com/company/eatbetter-app, founder linkedin.com/in/muratyasarr
- Complaints: www.sikayetvar.com/eatbetter-kalori-diyet-takip
- Apps: apps.apple.com/app/id6639614109, play.google.com/store/apps/details?id=com.aisparkslabs.eatbe…
- TikTok (TikHub search eatbetter kalori / eatbetter uygulama, 2026-11-11): @fit.kultur, @pt.samedcansengonul, @lifeofulya, @eatbetter.app
- Our databases: traffic (rank_day, crux_domain), appstore (app_profile, app_day, v_chart TR Health & Fitness Top Grossing), googleplay (app_day, app_profile, chart_rank), metaads/googleads/tiktokads (no records), 2026-11-11 queries