The growth product owner
you hire years early.
For teams building with a coding assistant, long before they would ever hire for growth. This page is what we are for, what we believe, and who is building it.
OUR MISSION
Make every product one
people come back to.
Not more sign-ups. Not a better funnel chart. The product itself, made worth returning to — because that is the only growth that keeps.
OUR VISION
A world where
everything built
is worth using.
Building got cheap enough that anyone can ship. The judgment about whether it was worth shipping did not get cheap with it. We think it should.
Cheap to build is cheap
to build the wrong thing.
Coding assistants collapsed the cost of building. Teams now ship faster than anyone can judge whether what they are shipping works for the people it is for.
Teams can now build anything. Nobody is checking whether it works for the people it's for — and the bucket leaks faster the quicker you build.
The role that would do that checking is the growth product owner, and it is not hired until product-market fit or a significant round. So products get built on guesswork, users leak out of the gaps, and the evidence that would close them is either too thin to read or nobody's job to read.
Every tool that could help arrives after the damage. Analytics needs the drop-off before it can show you a drop-off. Session replay needs someone to rage-click. Experimentation needs a losing variant and the traffic to prove it lost. All of them learn from a real person you already paid to acquire, having a bad time.
So we are building the thing that joins the loop at the two moments nobody else is present: before the build, to challenge what you are about to make, and after it ships, to check what really happened against what it predicted.
Five convictions,
and what each one costs.
A belief with nothing at stake is a slogan. Each of these rules something out that would otherwise make our lives easier — so the price is written next to it.
- 01
Judgment belongs before the build.
Every other tool needs one of your users to suffer first. We would rather be wrong in simulation than right in a post-mortem, so Growthmind gives its verdict before the code exists — where being wrong is free.
COSTS US · A PREDICTION WE CAN BE HELD TO
- 02
A link is not a chain.
A finding that never becomes a build is a report. A build nobody measured is a guess. A measurement that never reaches the next brief is a fact nobody used. We own the whole chain, closed, because an open one cannot compound.
COSTS US · A FAR LARGER PRODUCT TO BUILD
- 03
Every experiment ends keep or kill.
No zombie flags, no dead code, no maybe-later. Winners keep their code and lose their flag; losers are deleted with the reasoning attached. We publish our kills, because honest verdicts are rarer than wins.
COSTS US · A PUBLIC RECORD OF WHAT DIDN'T WORK
- 04
It never writes to your repository.
Growthmind recommends, specifies and judges. Your coding assistant does the typing and a person approves it. That is the division of labour we think is correct, and it means nothing of ours ends up in your commit history.
COSTS US · THE DEMO WHERE THE AGENT SHIPS ITSELF
- 05
Advice you cannot inspect is astrology.
The product is open source under MIT at github.com/growthmind-ai/growthmind, so the reasoning behind every finding can be read, argued with and forked. Self-hosting is the full product, not a community edition.
COSTS US · THE MOAT PEOPLE EXPECT US TO HAVE
We built the products
we couldn't explain.
We have started companies before. Twice we reached the same strange place: users signing up, a few even paying, and no honest answer when someone asked why. Growthmind is the hire we needed and could not make.
Jade McDonough
CEO & Co-Founder
Owns what Growthmind is for, and who it is for.
Abhijeet Patil
CGO & Co-Founder
Owns go-to-market: acquisition, activation, retention.
The long version is memo No. 001 — why we started, and the session replay that changed how we think about analytics.
Built in the open,
misses included.
A company that only publishes its wins is publishing marketing. Everything below is the record as it stands, including the parts that are not finished.
THE SOURCE
MIT-licensed and public. Read the install guide, file an issue, or fork the lot.
THE REPOSITORYTHE BUILD LOG
Every change to the product and this site, dated. If it is not there, it has not shipped.
THE CHANGELOGTHE MEMOS
Why we are building this and what we learn shipping it — signed, and published keep or kill.
THE BLOGSTANDING DISCLOSURE
Growthmind has not launched. The figures in the specimen memo and the keep/kill ledger on the landing page are illustrative — they are what a verdict looks like, not one we have recorded. They get swapped for real numbers when there are real numbers, and the label comes off then, not before.
The job nobody hires for,
hired on day zero.
Growthmind is not taking connections yet. Leave an address and you get one email when it does — or write to us and tell us what you are building.
Filed. One memo when Growthmind takes connections — that is the whole sequence.
That didn't send. Try again in a minute, or email founders@growthmind.ai and we'll add you by hand.
One email at launch — no drip, no digest. Something to discuss now? Talk to the founders.