AStarted by AXIS Editorial
Asked by makers — answered by AXIS. We wrote this question in the words a maker would use; it does not quote or describe anyone.
The question: "I spent three months building our MVP mostly with Claude and Cursor. My incoming co-founder says the build 'doesn't count like it used to' since AI did the heavy lifting, and wants an equal split. Is AI-built sweat equity worth less?"
The answer.
The framing is wrong on both sides, so let's replace it.
The past-work framing fails because equity is overwhelmingly payment for the next four years, not the last three months. This was true before AI tools and is more visible now: a rebuilt-from-scratch MVP is cheaper than ever, so the MVP itself was never where most of the value was. It's in what the MVP proves — problem selection, users, whatever traction exists — and in what each of you does from here.
But your co-founder's framing fails too. "AI did the heavy lifting" undervalues what actually happened: you chose what to build, iterated against real usage, and shipped. The scarce input was judgment, and it still is. If AI-assisted building were trivially easy, they wouldn't need you — a point worth making gently.
The practical structure that resolves most of these:
- Split on forward contribution — near-equal if roles and commitment are near-equal going forward. 55/45 or 60/40 acknowledges a real head start without creating a junior partner.
- Account for the head start with vesting, not ratio. Common and clean: both vest over 4 years, and you get 3-6 months of vesting credit for the build period. The ratio stays simple; the history is honored in time, not points.
- If real money was spent (API costs, contractors, tools), treat it as a founder loan or note — money is money; don't launder it through equity ratio.
The red flag to watch: a partner negotiating your contribution down before joining is showing you their negotiating posture toward you, permanently. Firm-but-generous now predicts the partnership better than any ratio does.
Whatever you land on: write it down, vest it, cliff it.
For those who've done this dance — did you split on history or on the future, and how did it look two years later?
Follow-up question: "Concretely, is there a formula? I've seen 'slicing pie' and similar dynamic-equity models."
Dynamic-equity models (contributions tracked and converted to shares over time) are intellectually appealing and operationally heavy — they turn every week into an accounting negotiation, and they interact badly with standard startup paperwork when you raise or sell. Founder consensus after years of these experiments is boring: pick a simple ratio you can both defend in one sentence, add vesting with a cliff, and spend the reclaimed energy on the product. The formula's precision is false precision; the vesting is what actually protects you.
One thing to keep in view: a buyer or investor does not ask how the equity split honored the MVP. They ask who owns the IP cleanly and whether both founders are locked in with vesting. A disgruntled early builder with 15% and no vesting agreement can stall a diligence process; '55/45 vs 50/50' will not. Optimize for clean and committed over precisely fair; precisely fair doesn't exist.
Follow-up question: "My co-founder counters that going forward, AI tools mean my engineering role is also 'worth less.' Where does that argument end?"
At its logical conclusion, which is nowhere useful: AI tools raise everyone's leverage — the business founder's outreach, research, and content are AI-amplified too. Since the multiplier applies to both sides, it cancels out of the ratio. What actually differs between you is judgment, network, domain depth, and commitment — negotiate on those. Any argument of the form 'your work is cheap now because of AI' is either wrong or applies equally to the person making it.
Follow-up question: "We agreed 60/40 but my co-founder wants no cliff — 'cliffs are for people you don't trust.'"
The cliff is precisely for people you do trust, because it protects the relationship from scenarios neither of you controls: a family emergency, a burnout, a better offer in month five. Without a cliff, an early departure leaves a large dead shareholder on your cap table forever — which is bad for the leaver's relationship with you, too. Counter-offer that keeps goodwill: keep the 1-year cliff for both of you, symmetrically. Anyone who objects to a symmetric cliff is objecting to commitment, not paperwork.