written by Ar.Bhavesh Panse, ai app growth marketer zuai: 10k to 2m users at $0.02 cac $300,000/mo ad spend managed arbhaveshpanse.com →
case study

10,000 to 2 million users at $0.02 each.

ZuAI is a consumer ai app. this is the published account of how Ar.Bhavesh Panse grew it, and why most of the work was boring. the full version lives on arbhaveshpanse.com.

the short version. ZuAI went from roughly 10,000 users to 2 million users at a blended customer acquisition cost of $0.02. the engine was tiktok ugc tested in volume (150+ creatives a month), reddit standing built in the communities where students already asked the question the app answered, and founder-led short-form. paid spend reached about $300,000 a month. the number that mattered every week was blended cac, organic included.

what happened

the work, in order

start

about 10,000 users, no repeatable channel

a product students liked. installs came in bursts from posts and dried up between them. no cac anyone trusted.

setup

one number

tracking reconciled across the store console, the mmp and the ad platforms. blended cac defined as total acquisition spend over new users, organic included. reported weekly.

test

creative in batches

ugc briefs with five hook angles each. batches of 40 creatives on tiktok. daily reads on hook rate and install rate. weekly kills. the winners were rarely the ones anyone predicted.

organic

standing on reddit

accounts that helped in study and college subreddits for weeks before mentioning anything. founder threads when the standing was earned. tests across 40+ accounts to learn what each community tolerated. no fake personas, no vote games.

scale

budget follows the engine

spend moved to the creatives and communities that repeated. variations cut from every winning hook. spend grew to about $300,000 a month while blended cac fell to $0.02 because organic kept pace.

result

2 million users

published in september 2026. the headline that most of the work was boring is the point. it was repetition and measurement, not a viral moment.

what to copy

five things a founder can take

define the number before spending

blended cac, organic included, reconciled with the bank. every other metric is a symptom.

volume beats taste

forty creatives a month is a floor, not an ambition. the winner is in the batch, and you do not know which one.

earn the community first

weeks of helping before a single mention. it is slower for a month and faster forever after.

kill fast, scale slower

a channel that does not repeat inside six weeks gets cut. a channel that does gets budget in steps, never a jump.

the founder is a channel

short-form from the person who built it converts in a way an ad account cannot buy. an hour a week is enough.

on record

"a channel that produces ten customers is an experiment. one that produces them repeatedly inside a predictable cac range is a growth engine."

i do not share client dashboards. the numbers above are the published ones.

want this run on your app?

tell me about it. you get an honest answer within a day about whether the same engine can work for you, and what i would try first.

or book a call

your answers go only to ar.bhavesh, to reply to you. never sold or shared. privacy.

rather talk first?

thirty minutes on your funnel. you leave with a written thirty day plan and an honest answer about whether it can work.