A 2,063-follower alt account hit 13.9M views: DeepFind sells "check him before the date," and makes money off TikTok alts and a weekly subscription

DeepFind is an app where you "enter a name, phone number, or photo and AI finds everything about that person online." It's only listed in the US, at $9.99 a week. It's not the first in this category, nor the biggest, and its reviews are bad. It got going because the team behind Sendit carried over a playbook wholesale: dozens of alts mass-posting "catch a cheater" skits, a five-star request on entry, and a hard paywall. Below: what it sells, where the traffic comes from, where the money comes from, and whether this path is worth following.

A 13.9M-view video

On Aug 18, 2026, an account called @user024562456 posted a TikTok video with only one line of caption, "yeah i had a feeling…", tagged #relationships #deepfindad #deepfindpartner. The account had 2,063 followers. By Nov 6, that video had 13.9M views and 475k likes. Three weeks later it posted "mac n cheese is DONE," which reached 8.51M views.

There are many more accounts like it: @randomgirl7734 (251 followers) posted "the app is deepfind yall" in Jul 26, reaching 5.05M views; @ally63471 (96 followers) wrote "turns out he was subscribed to other girls too... I found all his accounts on DeepFind" in Aug 5, reaching 410k views; @brittkneetailor (544 followers) posted "CHAT AM I COOKED? Deepfind exposed everything," reaching 3.25M views. All use a young woman's voice and all run "I looked him up and found..." skits, many tagged #deepfindad or #deepfindpartner, meaning they're ads or partnerships.

But their everyday output looks different. @user024562456 posted 5 videos over five days, Oct 9 to 8, each getting a few hundred to two thousand views (TikTok, 11-06). Dozens of alts post several times a day, most go nowhere, and a few lucky hits of a few million views carry the whole operation. That's DeepFind's growth engine.

What it is and who made it

The store listing is blunt: "Find Anyone Online with Ease." You enter a person's name, phone number, email, social handle, or photo, and it returns their public info and social profiles online, and you can follow up with questions in a chat. On opening the app, a window asks you to rate it five stars, followed by a paywall: $9.99 a week or $39.99 a month.

The developer is Orbit Intelligence, based in Culver City, Los Angeles, founded in 2025, backed by TenOneTen Ventures. Its core business is a people-lookup data API for AI agents (its site says it covers 37.9M people). DeepFind is the consumer version of that API, and the terms of service still use the name "Orbit." CEO Hunter Rice founded Sendit, the anonymous Q&A app that spread virally among teens via Snapchat and TikTok before selling a $9.99-a-week membership. He made Forbes 30 Under 30.

Some background to keep in mind: in September 2025, the US FTC and DOJ sued Sendit's parent company, Iconic Hearts, and Hunter Rice personally, alleging violations of children's privacy law, use of fake messages to lure subscriptions, and misleading charges (ftc.gov, justice.gov). We haven't found how that case was resolved.

Where the volume comes from

On Android, there was a concentrated surge in October, and it's fading now. The main battlefield is actually iOS: it was listed there over a month earlier, has 6,000+ US App Store ratings, 40% more than on Play, and still adds about 50 ratings a day between 10-28 and 11-06 (App Store US, 11-06).

Traffic comes from three sources:

• TikTok alt matrix, confirmed, the main channel. These are the dozens of accounts with a few dozen to two thousand followers described above. Some of the same accounts also tag #orbitpartner to promote the company's other product, which suggests a centrally run creator program where content and accounts are reused across products. There's also an iOS one-star review from Nov 1 saying "I keep seeing AI videos about it", so some footage may be AI-generated (unverified).
• Paid ads, possibly. A Play review from Oct 4 says "you guys are begging me to use this app in your ads." Whether the October spike on Play came from paid buys or a viral video is not visible from the outside. Those same days, @user024562456's new videos only got a few hundred views, which doesn't look like they were driving it (inference).
• Riding the category name, inferred. "DeepFind AI Search Agent" differs from category leader "Deepsearch AI Search Assistant" by a single word. By late 2025, TikTok already had videos like "I hid in the bathroom and looked him up on Deep Search" with 2.9M views, so people searching "deep search" would run into it.

Where the money comes from

Listed only in the US, it ranked #117 to #136 on the Google Play US Lifestyle Top Grossing chart (10-29 to 11-06). The weekly subscription is $9.99, an order of magnitude more than category leader Deepsearch AI's $1.99/month minimum. Actual revenue hasn't been disclosed, so we can't estimate it.

They charge hard, and the reviews reflect it. The store rating is 4.7, but that score is propped up: on Play, the number of ratings is 9.3% of installs, where a normal app sits around 1%–3%. An iOS review from Aug 15 says it plainly: "You have to rate it five stars before you even get into the app, then right away it wants you to pay, 10 dollars a week or 39 dollars a month." Of the 32 readable US iOS text reviews, 17 are one-star, and nearly all of them are about three things:

• It finds nothing: "it gave me nothing but a name and age" (6-21), "I put in my own info and nothing came up" (8-15), "Google works better" (10-07).
• Unexpected charges: "it said it was free for now, then charged my card" (10-27), "I chose the free trial and still got charged $17" (8-13).
• Hard to cancel: "there's no cancel option anywhere in the app, you have to go through the App Store to get out" (10-16).

There are five-star reviews too, and they describe the scenario where it actually connects: "I deepfinded him before the date, found some crazy stuff, and just didn't go" (6-13), "I was meeting someone I met on Facebook Marketplace and was scared, looked them up and felt way better" (6-16).

So the differentiator it claims, a proprietary people dataset, users don't feel. It got going on distribution and a paywall funnel, not on the product.

Who it's competing with

The "AI people search" category was built by Turkish company HubX with Deepsearch AI, and that's no longer a secret.

All DeepFind does is take the low-priced feature the leader sells in Latin America, repackage it as "catch a cheater," and sell it at a high price in the US, where the leader hasn't pushed hard.

If it were you

Verdict: C. Don't build another "Deep + Search" people-finder app. The category leader's revenue has already been publicly broken down, the space is packed with a dozen-plus reskins, the products mostly return nothing, and they run on paywall funnels where one-star reviews and refunds are the norm. DeepFind got off the ground because of a team that built Sendit and VC money. What's actually worth copying is its TikTok alt matrix, and that playbook can be used to push your own product.

Who it's for: small teams with an existing consumer subscription app, a target audience of women 18–25, and strong emotional scenarios (dating, relationships, friend gossip, outfit roasts). If you can organize a batch of regular people or batch-shoot short skits yourself, this playbook is cheapest to learn.

Who it's not for: anyone wanting to build people lookup, background checks, or face search. You won't get cheaper traffic than HubX or better data than Whitepages. In the US this category also brings in FCRA (background-check rules), privacy, and platform review, and Hunter Rice's previous company was sued by the FTC and DOJ over its charging practices. It's also a bad fit for anyone hoping the word "AI" alone will differentiate them, since ChatGPT can look people up for free.

How to do it: if you must touch this need, pick one angle nobody does well: verifying dating matches (is the photo really them, any record of being married), and it has to actually surface something. Hook up a licensed public-records or reverse-image data source, charge per check as a one-time purchase, no weekly subscription, no hidden cancel button. Flip the three complaints in the bad reviews into your product requirements: it finds things, it doesn't charge you secretly, and you can get a refund.

How to get users: copy DeepFind's approach and run 20–50 new accounts, each with a fixed persona (just broke up, dating someone new, checking up on a friend's date), posting 2–3 videos a day using one template with swapped plots: "I looked him up before the date, and then..." Tag #brandpartner for ad disclosure. Most will get a few hundred views. Just one or two per month breaking a million pays for the whole thing. Keep the accounts that produce hits and reuse them for your next product.

When to move: there's no rush to learn the playbook. If you do touch the people-finder category, the peak for dating topics is before Valentine's Day 2027 (February). After that, interest and the window both fall off.

What it would mean we got it wrong: ① DeepFind's iOS US-store rating growth speeds up from about 50 a day to several hundred a day while its one-star share falls. That means the product is really getting better, and this category still has room at high prices in the US. ② HubX or another player takes the US market with low prices and multiple languages, and DeepFind drops off the Top Grossing chart. That means the US high-price gap is closed too. ③ Conversely, if the App Store or Google Play starts pulling people-finder apps in bulk, or the FTC's case against Sendit ends in limits on how this kind of paid model works, this category is off the table entirely.

---

Data and sources: Google Play store page, install counts and US Lifestyle Top Grossing chart (2026-07 to 11-06); App Store US-store ratings and reviews (id 6775974979,11-06); public TikTok videos and view counts (searched deepfind and #deepfindpartner, pulled 11-06); deepfindpeople.com terms of service, orbitsearch.com, developer.orbitsearch.com; TenOneTen investment page; FTC and DOJ indictment notices on Iconic Hearts (2025-09); 36kr's report on Deepsearch AI (citing Diandian data); AppGazers. Ad channels, the October scale-up cause, and the AI-generated creatives are inferred or unverified. Revenue is not public.