cofounderAutopilot for your entire marketing team.
It researches what wins in your niche, makes the creatives, posts them, reads the numbers and doubles down — daily, without being asked.
Made last night, by itself
Eighteen of these, before you woke up.
All in the customer's own kit — their ink, their tan, their cream. Three are variants of the format already winning; two chase a hook nine competitors are running; two are cheap tests it will kill by Tuesday.
Meta · winning format · 4.9 ROAS
Meta · transparency · 3.4 ROAS
Organic · objection handling
Meta · price-per-cup · testing
See the full set of eight — with the reasoning behind each one.
Why you're here
Marketing is the job you never signed up for.
You started this to build the thing. Instead you're the media buyer, the copywriter, the editor and the person who forgot to post for eleven days.
You post when you remember.
Which is the weeks you aren't shipping, hiring or on support. Nothing compounds, so every month starts from zero.
Every creative needs a brief.
An hour of art direction you don't have, then a week of waiting, for one asset you'll test once and never learn from.
You can't tell what worked.
Three ads is not a test. With no volume there's no signal, so you end up guessing and calling it strategy.
An agency wants $4k and a quarter.
Retainer, kickoff call, strategy deck. Ninety days before anything is live, and they still ask you for the assets.
Meanwhile the ads beating you are the same twelve formats, running untouched for sixty days.
That's the whole game: find the formats that work in your niche, make a lot of them, kill the rest fast, and never stop. It isn't clever — it's just more work than one founder can do between everything else.
So don't do it. Hire the loop.
It doesn't answer prompts. It runs a loop.
Four moves, repeating. Every turn is informed by the last, so week six outperforms week one with nothing new from you.
Research the niche
Pulls the ads running against you, the hooks they lead with, and how long each has been live. The ones still spending are the ones that work.
Make the creative
Writes the angle, art-directs the frame, produces static and UGC-style video in your brand kit. Fifteen variants, not one.
Publish & spend
Builds the campaign, splits the budget, schedules the organic posts, ships. Your connected accounts, real placements.
Read & iterate
Kills losers by morning, clones the winning format into new angles, and tells you in one paragraph what it learned.
The core four
Paid
Creative testing at a cadence no freelancer will match: fifteen variants a night, budget shifted to whatever survives the morning, losers cut before they spend twice.
Control
You stay the creative director.
It does the labour, not the deciding. Set the leash per channel and change your mind any day.
Draft only
It researches and makes; nothing leaves the app until you say so. Good for week one.
Approve firstDefault
One notification a day with the batch. Thumb up, thumb down, or edit the line. It posts what you cleared.
Autopilot
Inside a budget cap and your brand rules it ships alone and reports after. Most founders land here by month two.
Branding
It knows what you look like before it makes a single ad.
Start with a name and a sentence: it builds the wordmark, palette, type pairing, photography direction and a voice guide it then obeys in every caption. Already branded? Upload the kit and it adopts it.
The obvious objection
“You want me to give an AI my ad account and my card.”
Correct, and you should be uneasy about it. So the guardrails aren't a promise in a support doc — they're the four things below, and you can check every one of them from inside the app.
A hard cap, not a target
You set the monthly ceiling. It throttles itself days early to stay under, and tells you when it does. It has never got the authority to exceed it — not a rule it follows, a number it can't pass.
Brand rules it can't break
Words it will never use, claims it can't make, colours it won't leave. Written on day one, enforced on every asset before it's queued — not checked after it's live.
Everything is reversible
Every post can be pulled, every campaign paused, every decision undone from the notification that announced it. Revoke access on Meta's side and it stops the same minute — you don't have to ask us.
A log of every action
Timestamped: what it did, what it spent, and the reasoning it had at the time. Not a feed of activity — an auditable record you could hand to an accountant.
And on week one it can't publish anything at all. Draft-only is the default — it earns the leash by showing you work you'd have approved anyway.
See the three modesProof, such as it is
We have no customers yet. So it markets us.
Every ad that brought you to this page was made, published and iterated by the product, against our own budget. The numbers below are ours, updated nightly, including the bad weeks.
It's the only honest proof a pre-launch product can offer: not a testimonial, just our own account left open.
Our CAC
$31
from $88 in week one
Creatives
2,140
none touched by a human
Signups
610
waitlist, 11 weeks
Bad weeks
3
all on the dashboard
Week 4, in its own words
“Everything I made this week lost. Six angles, all under 0.7% click-through. I've paused them, put spend back on the format that works, and I'm testing a different hook tonight.”
We left this on the public dashboard. A product that only shows its good weeks is a product you should assume has bad ones.
Once there are customers, this section becomes their numbers. Until then it's ours, and you can watch it work in real time.
Pricing
What this replaces.
Not a tool you add to the stack. A line item you delete from it.
The agency
$4,000/mo
Three-month minimum, kickoff deck, and they still ask you for the assets.
The freelancer
$1,200/mo
Eight creatives a month if you write eight briefs. You won't.
Doing it yourself
6 hrs/wk
The six hours you'd rather spend on the product — and it still slips.
Cofounder
$49/mo
Plus your own ad spend, capped by you. Roughly 18 creatives a night.
Everything, one price
The honest maths
If it finds one format that beats your current best by 20%, it has paid for itself twice over on a $1,200 spend.
And if it doesn't, you've lost $49 and a fortnight — not a quarter and a retainer.
Hire the cofounder you can't afford yet.
14 days free · cancel in one click · your ad accounts stay yours