Four weeks with only 103 installs, then thousands a day in week five: the ad-spend ledger behind an AR tracing app

PrismArt is an Android AR tracing app: you prop your phone over paper, a semi-transparent image appears on screen, and you trace the lines. It hit 160k installs in just over three months and reached #25 on the Art & Design free chart in Vietnam. But it didn't grow organically. Behind it are a Vietnamese publishing company, a 10M-dollar user-acquisition loan, a set of developer accounts, and a category already crowded with 88 similar apps. Below is how it got going, where the money came from, and which parts you can actually use.

The day it took off, Aug 24

PrismArt launched on Google Play on Jul 2026, 26. For nearly four weeks almost nobody installed it: by Aug 21 total installs were only 103 (third-party estimates, same below).

Then on Aug 24 installs suddenly picked up, reaching over a thousand a day. From mid-September to October the daily average was 2,000–3,000; from late October to early November it was 4,000–7,000. As of Nov 6, Google Play showed 164,417 total installs, with 7 added in the last 37,660 days.

Organic growth doesn't look like this. Organic growth usually comes from a video that spreads, rises, then slowly fades. PrismArt went "zero, zero, zero, then suddenly open, then a steady daily climb." That's what someone starting to spend money on a given day looks like (inference, high confidence).

Around Sep 8, a Southeast Asian fundraising outlet reported that Vietnamese company Tevo had received a 10M-dollar non-dilutive user-acquisition loan from Singapore's PvX Partners, secured against 45 of its apps and nearly 200M installs (DealStreetAsia, ce-media.asia, and e27 citing it). PvX's model is to fund the user purchases up front, then take a share of the revenue those users generate.

PrismArt's developer website is falcon-mobile.tevo.app.

What it is

The store listing lists five features: AR camera projection (project the image onto paper, canvas, even wood), one-tap photo-to-line-art, hundreds of templates (anime, animals, portraits, architecture, lettering), time-lapse recording, and opacity and edge sharpness controls. The instructions are blunt: "Place your phone on something steady, a glass sheet or tripod, pick an image, aim it at your paper, pick up a pencil and trace."

The top apps in this category all have these five features. The description is stuffed with search terms like ar drawing, ar tracing, trace drawing, sketch app, anime drawing, learn to draw, and best drawing apps, and the package name is itself a keyword string: ar.draw.sketch.trace.paint.projector.

Money comes from two places: ads inside the app, and one-time in-app purchases at $1.99–$3.99. It doesn't appear on any top-grossing chart, so revenue is probably mostly ads (inference). Its ad accounts list three direct networks, Google AdMob, Meta Audience Network, and AppLovin, plus a long tail of resellers.

Who publishes it

Tevo's website describes itself as making "mobile and AI products people use every day, 200M+ users." Its partners page spells out the first type of deal clearly:

You build the app. Tevo runs user acquisition and monetization.

"You build the app. We handle acquisition and monetization."

PrismArt's developer account is named Falcons Mobile App, with a generic gmail contact. Following the "developer site sits under tevo.app" thread, at least 8 developer accounts like this show up on Google Play. Each uses a different gmail and runs one app in a category already proven to work: Live Earth Map (15.8M installs), ID Caller (11.8M), Natal Chart (4.04M), Dramini (short dramas), browsers, dating apps, and so on. The Falcons Mobile App account itself, besides PrismArt, also has a horoscope app called Zodiakos with 565k installs, gaining 100k in a week.

So PrismArt is one company's piece on the "AR tracing" square. From the outside, you can't tell whether Tevo built it or an outside developer handed it to Tevo to publish.

Where the money went

Only cheap countries. PrismArt has only charted on free Art & Design charts, all in emerging markets (Google Play charts, 11-06):

In these countries an install is cheap and the chart bar is low: a few thousand installs a day can get you into the top 30 of Vietnam's category chart, and charting brings more organic search on top. As long as one user's ad revenue exceeds what you paid for them, the machine keeps running. Which channels, what each install cost, and whether it pays back are all invisible from outside.

It has almost no social presence. Searching "PrismArt" on TikTok and X turns up nothing (11-06), and no media has covered it. The store listing shows no rating and no visible reviews.

How crowded the category is

AR tracing has been a major traffic source in Google Play's Art & Design category since 2023. Google Play has 88 apps with AR draw, trace, or sketch in their name or package, with 420M cumulative installs combined and 5.94M new in the last 7 days (11-06). 2026 such apps launched after May 24. PrismArt's 37.7k per week is 0.6% of the category's weekly new installs.

(Google Play install counts, 11-06.) ArtHub launched two months after PrismArt and already has over five times its installs. Newer launches through 9–October include Arteasy, Mintexecto's AR Draw, Glowtrace, Tracey, SketchSpark, and more. About ten AR tracing apps sit in the top 30 of Vietnam's Art & Design free chart.

The demand is real, and TikTok is what fuels it. Searching "ar drawing app" on TikTok, the top results look like this: an account called ar_drawing.anime posted "I can't draw anything until I met this ar drawing app" in May 14, 1.2M views; newbie_drawing1's "Trace and draw anything with AR Drawing app" has 510k views; an account named pictrace_ardrawing posted five or six "Free AR Drawing app" videos in a single day on Nov 4, one with 50k views (TikTok, 11-06). From the names, most of these look like the apps' own official or partner accounts (inference). There's also a #ardrawingweb hashtag for web-based tracing that needs no install, with individual videos from 150k to 2M views.

In short, the category's organic traffic lives on TikTok, and rival creators who post every day have taken it. PrismArt has none of it. Its traffic is bought.

Why it could take off

Three things, none of them in the product:

1. Demand already exists: a category adding 6M new installs a week, so you don't have to educate users.
2. Money to keep buying: behind it are the ad and monetization operations of 45 apps, plus a loan earmarked for user acquisition.
3. Only cheap markets: in Vietnam, Mexico, and Nigeria, a few thousand installs a day can put you on a category chart.

Almost no differentiation. What sets it apart from the leaders isn't in the app, it's in the publishing operation behind it: the money, the account network, and the ad monetization.

If it were you

C. You can learn a playbook here, but don't copy this app. AR tracing already has 88 apps, with the leader at 60M installs, and new copycats keep launching every month. PrismArt's 0.6% share comes from loans and a full publishing operation you can't access alone.

Who it suits: small teams already buying Google Play installs who can work out cost per install and per-user ad revenue to two decimal places, or people who already run several ad-monetized apps and have data to bring to a user-acquisition loan negotiation.

Who it doesn't suit: anyone who hasn't bought traffic before and expects an app to take off through ASO and organic installs, or anyone hoping to break into this category by making a "more accurate, easier-to-use" tracing app. The features have long been built.

How to do it: what you can take is the method, not the category. First, pick the square: find a proven tool category where ads can monetize but the market isn't saturated yet (check how many apps are on the category chart, how big the leader is, and how many were added in the last three months), rather than squeezing into AR tracing. Second, pick the market: cold-start in countries like Vietnam, Mexico, or Nigeria, where installs are cheap and the chart bar is low, and use a very small budget to test whether one user's acquisition cost can be covered by ad revenue. Only add spend once it breaks even. Third, replicate: one account per square, and rerun what worked in a new category. Once you're at scale, user-acquisition loans like PvX's, secured against your existing app data, let you grow spend without selling equity.

How to get users: paid acquisition is the core. At minimum, run one of Google App Campaigns, AppLovin, or Meta well (I can't verify which PrismArt used). If you do want a visual tracing app, the TikTok official account is the cheapest supplement: a peer's "I couldn't draw until I found this app" clip can rack up millions of views, and PrismArt skipped that entirely.

Timing: this playbook has no fixed window. It works whenever a new tool category emerges. But for AR tracing specifically, the window has already passed.

What would prove this read wrong: (1) PrismArt keeps growing weekly installs without new money and pushes into the top 10 of Vietnam's category chart, which would mean the category is nowhere near saturated in emerging markets and late entrants still have room, so the grade should go up; (2) it starts charting in expensive markets like the US and UK, which would mean its monetization is stronger than it looks and in-app purchases sell well beyond ads; (3) conversely, if its daily installs suddenly drop to a few hundred after December, that would suggest the spend stopped because it no longer paid back, and the playbook has stopped working in this category.

Data and sources: Google Play install counts and the Art & Design category chart (10-29 to 11-06); daily installs since launch are third-party estimates; PrismArt store listing copy (11-06); falcon-mobile.tevo.app/app-ads.txt (11-06); the tevo.app homepage and /partners page; ce-media.asia 2026-09-09, e27 2026-09-08, and undiluted.com reporting on PvX and Tevo; public TikTok videos and view counts for searches on "ar drawing app" and "PrismArt" (11-06). That PrismArt's growth comes from paid acquisition is an inference drawn from the install curve, the publisher's public business, and the loan's stated use. The specific channels and costs are unverified.