We back founders hunting the thing that matters.
Maybe you have found the thing. Maybe you are still hunting it. What we look for is relentlessness: founders swinging at how an entire industry works over the next 20 years, instead of adding a feature to how it works today.
We move fast on decisions, say what we think, and a no comes with reasons you can act on.
BACKED BY BUILDERS FROM HUBSPOT, SHOPIFY, DROPBOX, VIMEO, BOLT, OPENAI, ZENDESK, SQUARE, REVOLUT, VINTED, NORDVPN, DEEL, TWILIO, VERIFF AND WISE.
Four things we look for.
We will be flexible on almost everything else. On these we are not, because they are what separates a company that lasts from a company that ships.
Something you know that the market has not caught up to yet, usually older than the company itself. You should be able to name the part of it that gets laughed out of the room.
NOT: A ONE PERCENT TWIST ON SOMETHING ALREADY OBVIOUS.
You are going at how an entire industry works, even when the first move is deliberately narrow. Stripe's wedge was seven lines of code to take a payment. The wedge is where you stand, not where you are going.
NOT: A BETTER VERSION OF A THING THAT ALREADY EXISTS.
The foundation model layer is already largely captured. The layer built on top of it has barely started, and that is where we invest. Software, not hardware.
NOT: COMPETING WITH THE MODEL LABS, OR WITH NVIDIA.
Agents in your loop, and learning measured in days. How many real hypotheses you test in a week is the second scarce thing after the belief itself.
NOT: AI IN THE PITCH AND A QUARTERLY RELEASE CYCLE.
It does not have to arrive fully formed. Some teams are early, and a belief can sharpen through real contact with the problem: Twitter fell out of a podcasting company. What it cannot be is borrowed, safe, or missing. Stay adaptable on the solution. Hold the line on the bet.
Send the deck as it is.
You do not need to polish it or write a cover note first. The team that reads every deck reads yours, and you get a yes or no in five business days, with the reason.
NOT READY TO PITCH YET?
PRESSURE-TEST YOUR PITCH INSTEAD
Drop the deck and it gets read the way our deal leads read it: built on the hundreds of companies they have backed, and the thousands they have not.
The best GTM operators alive, inside your company.
Once you have found the thing worth building, moving is the only thing that matters. So we put an operator who has already run your exact function, outbound sales, PLG, or marketing-led growth, at Zendesk, Square, Scale AI or OutSystems, inside your company. Every company we fund gets the same operators, on a schedule, never a one-off intro.
Built Zendesk's SMB-to-enterprise motion from scratch, through IPO and a $10.2B acquisition. Then led sales at Xero and Prezi.
Square's first sales hire. Built the sales org from one person to 200, then founded and sold his own company, Ramped.
Former Head of Growth at Scale AI. At Braid, zero to $1M monthly volume while cutting acquisition cost from $200 to $40.
Built PLG from zero at OutSystems, then 3x'd ARR at Softr in a single year.
Build things that matter. Then throw everything you have at it.
Everyone thinks it got easier. Cheaper to build, faster to ship, no gatekeepers left. But to build something that lasts, that wins a market, that builds its own gravity? Harder than it has ever been. When building goes to zero, building stops being the thing that sets you apart. Two things are left: conviction in a problem worth obsessing over, and the speed at which you learn.
Conviction cannot be borrowed. We are all thinking with the same machine now, the same models, trained on the same averaged-out record of everything anyone has ever said. Ask it what to believe and it hands you the median, and the median is everyone's, and everyone's is worth nothing. An LLM can sharpen your conviction. It cannot build it for you.
Great founders still die at seed. They rarely die on conviction or product. They die on go-to-market, on which experiments they run and in what order. That is the one place we go all in: the best GTM operators in the world on your problem, turning your read of the customer into the right experiment in the right sequence, in days rather than quarters. You bring the conviction. The average is free, and worthless. So find the thing you have to do, and run like hell at it.
Read the digest.
ALL ARTICLES →Each article takes one go-to-market problem we hit in the portfolio, shows the working, and ships with a tool you can run on your own numbers.