Satellite treasure hunting: a Turkish app turns folk "define" into a land report for 559 lira per report

Turkey has a whole crowd of people who spend their weekends on hills with metal detectors hunting for ancient treasure. They call it "define." This September, a developer named Kasım Yılmaz shipped "Arazi Tarama" (Land Scan): draw a box on a map, get a report in about a minute on "mineral traces, groundwater, buried structures, cave voids." The first analysis is free, but the exact spots are locked, and unlocking costs 559 lira. The Android version passed about 15k installs in under two months and sits near the top of the Turkey Tools Top Grossing chart, around #30. Below: what it actually sells, where its users come from, why it can take money, and whether you can run the same play, and how.

A one-star review saying "found nothing"

On Oct 2026 28, a Turkish user left a one-star review on the App Store titled "Tamamen para tuzağı" (a total money trap):

"We ran a land analysis, the map showed valuable minerals, structures and water, but when we went out to search, we found nothing."

The next day another one-star: "How can you possibly see valuable minerals and water from a satellite? Don't fall for this trap." The same day, a five-star: "Gerçekten işe yarıyor" (It really works).

Put these three reviews together and you have most of the app: users really did take the reports out to dig, and they really did pay. A product that makes money off "there might be treasure under your land" can get polarized reviews and still have steady revenue.

What it is

The store listing is restrained: "An analysis app that combines satellite, geological and elevation data for a plot of land and interprets it with AI." You draw a boundary on the map, or enter an address or coordinates, and about a minute later you get findings: numbered points on the map, each with a score and a Yes / Possible / Weak signal rating.

The official site lists the kinds of findings: traces of valuable minerals, groundwater, "buried structure traces," "void / cave traces," plus some that sound more respectable: what to plant, distance to the nearest active fault, whether you can install solar panels. Results can also be turned into 3D terrain that you can rotate, compare across seasons and 2014 years of satellite imagery, and export as a rotating video in one tap.

Beyond the report there are three paid add-ons:

• Phone metal detector: uses the phone's magnetometer, "go to the spot the satellite points to and scan with your phone." The site itself says detection range is only a few centimeters for small objects and 10–30 cm for larger ones.
• Symbol interpretation: photograph carvings on a stone, the AI tells you what they traditionally mean, then suggests which nearby plots to scan. In the treasure-hunting community this is called "işaret" (signs), an old define topic.
• Expert opinion: send your results to an "experienced field expert" and get a reply via push notification (about €6.99).

Prices are one-time credits, not a subscription (per the site, 11-01):

On Google Play the in-app purchase range is $2.99–99.99 per item (Google Play, 11-06).

All data sources are public: ESA Sentinel-2, Copernicus 30-meter elevation, Esri satellite imagery, the USGS earthquake catalog, geological maps, with interpretation by Google Gemini. The site has a section titled "We are not a magic wand," and a prominent notice that treasure digging in Turkey requires a permit from the museum directorate under Law No. 2863, and mineral prospecting requires an MAPEG license. The app cannot replace any permit.

Metric

The two platforms are far apart: Android has about 15k installs, while iOS has only 14 ratings. iOS came after Android was already working. Chart position isn't a matter of headcount either: one 10-credit pack is a bit over $100, and on category charts like Tools and Utilities, a few big purchases are enough to climb. Revenue isn't public, so the actual figure can't be seen from outside.

People were already doing this business

"Finding treasure by satellite" (uydudan define bulma) is not new in Turkey:

One more important layer: for years, blogs run by Turkish detector sellers have been writing that "satellite treasure hunting is a scam." The demand, the supply and the backlash all already exist. What was missing was a phone app that installs easily and looks like a legitimate tech product. Arazi Tarama filled that slot.

Where users come from

Almost all of it is paid acquisition. Its privacy policy is clear: it integrates Meta and TikTok SDKs and sends install, open and purchase events, with amounts, server-side back to Meta to measure ad performance. Its Facebook page "Arazitaramax" posts standard ad creatives, and one caption reads:

"Even grandpas have figured it out! Son, stop wasting your time with picks and shovels. Download Arazi Tarama."

I couldn't find how many ads were run or how much was spent.

Its own content accounts barely worked. The TikTok account @arazitarama has 79 followers, with 173–417 views per video. In Oct 5, a small account with 483 followers, @istanbul_kelebek, posted "Analyzing the Göbekli Tepe stone circle with Arazi Tarama," which got 524 views and included an Android link. Likely a paid partnership.

App store search brought in a little. The Android title reads "Arazi Tarama: Land Scan AI"; searching "define" in the Turkey iOS store, it ranks first.

Android installs rise steadily by several hundred to a bit over a thousand per day, with no spikes and no clear tie to any video. That looks like daily-budget performance ads, not a viral hit (inference).

Why it can make money

This crowd already spends big. An underground imaging device costs tens of thousands of lira, so a 559-lira analysis looks like small change to them. The emotional pull is strong: "there might be something buried under my land."

Neutral name, compliant copy. It's called "Land Scan," not "Treasure Finder." The data sources are listed one by one, the legal disclaimer is prominent, and legitimate uses like farming, earthquakes and solar are included. That helps it pass store review and Meta's ad review, and makes users trust it more.

The paywall sits right on the itch. The free analysis still shows 3D and still tells you "there's a signal here," just not where. Once users see a locked pin on the map, paying is the easiest next step.

One impulse, sold in pieces. Satellite scan → verify on site with a phone detector → photograph stone symbols for AI interpretation → consult an "expert" → join the community. Each step rides the same urge.

Fast iteration, and each change is ad material. Within 10 days of the iOS launch it went from 2.1.8 to 3D terrain, seasonal comparisons, ten-year-old satellite imagery, and finds of caves and underground cities. Each one is built to make a viewer go "wow" in an ad.

There's no technical barrier: public satellite data, elevation and Gemini can be stitched together by anyone. What it really got ahead on was proving the combination of Turkish treasure-hunting demand, Meta/TikTok paid acquisition and high-priced one-time credits.

Crowded or not

Pretty empty. I couldn't find a Turkish-language "satellite + AI one-tap report" app copying this one. Define Rota 3D launched the same day. That's a parallel race, not copycats, and they're still shipping the old-school file formats. No press coverage, no X discussion, no disclosed revenue. Only people in Turkey's treasure-hunting scene who've seen the ads know it exists.

But the pushback has started. The App Store has 3 one-star reviews calling it a paid trap. And @kayitliantikarsivler, an 17 account with 8,900 followers on TikTok, posts videos warning people about "AI + satellite treasure" scams.

If it were you

Verdict: the playbook is clear and copyable, and the markets outside Turkey for treasure and water finding are still open. But this is a gray-edge business that lives on paid acquisition. The reputation and platform risks are real.

Who it's for: small teams that can run Meta / TikTok performance ads, optimize on paid-event postbacks, and understand the language and subculture of one local market.

Who it's not for: anyone counting on SEO or organic content to grow (their own content account never took off), or anyone who can't stomach selling a probability report where many buyers find nothing.

How to do it:

• Switch markets rather than fight in Turkey. Arabic-speaking treasure circles in the Middle East and North Africa, the Balkans, and Iran all have similar buried-treasure folklore and detector businesses. They've already added Arabic and German (for Turkish-origin Germans), so they see it too. Get there before they do.
• Switch the need. Water finding and well drilling (in rural areas, people hire someone to "read the water" before digging, an old trade), gold prospecting, and feng shui site selection are all cases where people already pay for the outcome but have no decent app. They're also cleaner than digging up artifacts.
• Copy the funnel: first scan free, results shown but the key info locked, one-time credit packs instead of subscriptions, and a few small paid add-ons beyond the scan. Re-scanning the same plot shouldn't cost extra; that small kindness cuts down on "scam" complaints.
• Do the compliance work up front: neutral naming, data sources and "this is a probability estimate" on the first screen, and clear local legal disclaimers. That's what gets you through app review and keeps your ad accounts from getting banned.

How to get users: Meta and TikTok performance ads, server-side purchase postbacks, optimize for paid users. Creative is 3D terrain rotation videos and a locked glowing point on a map, plus a catchphrase locals will instantly recognize. Organic as a supplement: win the store title and keywords for local-language terms like "treasure search" and "land scan."

When to move: within three to six months. They're already expanding into Arabic, and this stack can be rebuilt in two weeks, so the window won't stay open long.

What would mean you got it wrong: their Android version gets pulled, or their main ad account gets banned; daily Android installs drop to a couple hundred, or they fall out of the top 100 in Turkey's Tools top grossing chart; or, when you run your first test budget in a new market, your cost to acquire a paying user comes out higher than the price of one analysis (say, over $15 per payer while the 10-credit pack sells no one). That means this audience isn't willing to spend in the new market, or the ads aren't getting approved.