AI app marketing strategy: the CFCP framework, written down
An AI app marketing strategy you can run weekly: community, founder, content and product led growth, one loop across all four, and the order to start them in.
most AI app marketing strategy documents are a channel list with a budget next to each line. they read well and they change nothing, because a strategy for an AI app has to survive the category moving every month. the only thing that survives that is a loop, not a list.
this is the framework i run for every consumer AI app, written out so you can run it without me. it is called CFCP because every consumer app grows through four engines, and most founders are running one. the AI app growth marketer page is the hire version of this; this post is the do-it-yourself version.
The four engines
C, community led growth. your buyers are already talking, on reddit and X, about the problem you solve. reading that is the cheapest research you will ever do and the only channel that tells you why anyone cares. building real standing there, so a mention is welcome instead of removed, gets you your first 10,000 users. it cost me 40+ banned accounts to learn how; the SAFE Reddit playbook is the result.
F, founder led growth. you are the least copyable asset the company has and usually the least used one. phone-shot video on a cadence you can keep, the lessons you already learned written as posts, and outreach that runs without you living in the inbox. founders who post consistently cut their paid CAC, because the ads land on a warm name.
C, content led growth. short form in whatever format the platform is rewarding this month, at volume. thirty videos to find two. the winner at ZuAI was shot by a college junior on a phone, did 2.1 million views and 120,000 users, and is still the best ad in that account. a $2,000 creator video did 12,000 views. volume beats polish.
P, product led growth. acquisition is wasted if week one loses them. onboarding that gives a person a win before it asks for anything, A/B landing pages because the page is half the ad, and conversion work on the step that is actually leaking, not the one that is easiest to change.
One loop across all four
every engine runs the same four steps, every week:
analyze. read what the numbers already say before touching anything. the ad account history, the retention curve, the creative graveyard. most graveyards hold a winner nobody scaled.
execute. ship the batch. creative, posts, pages, tests. execution starts in week one, not week six.
iterate. winners get budget and variants. losers die without ceremony. at ZuAI, every sunday the new batch went out and every monday the winners moved into the scaling campaign.
repeat. the same unglamorous week, again, until the number moves. the channel changes. the loop never does.
The order to start them in
do not start all four. start with the one that is obviously underused, prove it, then add the next.
if you have under 10,000 users: community first. reddit and X give you users and the exact language your customers use, which every later engine needs.
if retention is under 20 percent at day 7: product first, and nothing else. no engine fixes a bucket with a hole in it. a teen investing app i worked on hit 10,000 users at 5 percent day 7, and the fix was rebuilding the product into ZuAI, not another channel.
if you have a product people come back to and no repeatable channel: content, at volume, organic first. organic is the free testing ground. nothing gets paid budget until it has already earned attention for free.
if a channel already works: founder, because it lowers the CAC of everything else, and then paid, scaling only what organic proved.
What makes the strategy specific to AI apps
three things, and they are the reason a 2022 playbook does not transfer.
the buyer is still learning what the category is, so every piece of creative has to teach before it sells. the category moves monthly, so the loop is weekly and the quarterly plan is fiction. and the model is not the moat, because a competitor ships the same one next week; distribution and retention are the only defensible things you own.
write the four engines down. pick one. run the loop for four weeks. that is an AI app marketing strategy that survives contact with the market. the free calculators will tell you whether you are ready to start, including one built to say no.