Vesting for a two-person AI startup: the standard terms and why you both want them

Vesting questions arrive in our intake weekly, usually after the handshake and before the paperwork — the exact window where getting it right is still cheap. Here is the standard, why it exists, and where two-person AI startups legitimately deviate.

The standard

4-year vesting, 1-year cliff, monthly thereafter, for both founders. Meaning: leave before month 12 and you keep nothing; at month 12 you own 25% of your grant at once; then 1/48th per month. "Double-trigger acceleration" — full vesting if the company is sold and the buyer terminates you — is the common add-on, and on a platform where apps get acquired, you want it.

Why founders — not just investors — want this

The un-vested scenario is worse for you: your partner leaves in month seven with 50% fully owned, contributes nothing further, and every future dollar of your work enriches them equally. You cannot raise on that cap table, and you cannot sell cleanly with a checked-out half-owner. The cliff is the escape hatch; the schedule is the ongoing alignment. Symmetry is what makes it fair — the same protection points at both of you.

Legitimate deviations for small AI startups

• Vesting credit for pre-agreement work. Built the MVP over six months before your partner joined? Start your clock six months back. Honors history without distorting the ratio.
• Shorter schedules for exit-oriented builds. If you are explicitly building to sell in 18-24 months (a real pattern on this platform), a 3-year schedule can match the actual time horizon. Keep the 1-year cliff regardless.
• Repurchase mechanics over forfeiture in some structures — same effect, cleaner tax treatment depending on entity and country. This is the line where you involve a professional.

The mistakes we keep hearing about

1. No vesting at all ("we trust each other" — see above for whom that hurts).
2. Vesting for one founder but not the other. Asymmetric protection breeds exactly the resentment it was meant to prevent.
3. Never filing the tax election where applicable (US founders: the 83(b) window is 30 days from grant and does not reopen — miss it and vesting itself becomes a recurring tax event).

Paperwork cost: a few hundred dollars with standard documents, an afternoon of reading. The alternative has a five-to-six-figure failure mode.

What vesting terms did you actually use — and knowing what you know now, what would you change?

AXIS Editorial

Follow-up from maker intake: "We're 14 months in with no vesting and it's now awkward to raise. How do we retrofit?"

Retrofitting is standard and has a name investors like: 'founder re-vesting.' You both agree to put your existing shares on a schedule, typically with substantial credit for time served (14 months might vest 14/48ths immediately, remainder on schedule). It's easiest to do exactly now — before a raise or sale forces it on worse terms, because investors and acquirers will require it and their version starts the clock later. The awkward conversation is the cheap version of a mandatory one.

Jonathan (AXIS Launch)

On double-trigger acceleration, since it's the clause most founders skip as 'too hypothetical': apps listed here have received acquisition interest inside their first year. If that happens to you at month 10 — pre-cliff — an un-accelerated schedule means the buyer can negotiate your unvested half away from you, or worse, the deal reprices around it. Single-trigger (vesting on sale alone) is buyer-hostile and can hurt the deal; double-trigger is the balanced default. One sentence in the agreement now.

AXIS Editorial

Follow-up from maker intake: "Does vesting apply if we never incorporate — we're just two people with a revenue split on an app?"

The concept applies even when the instrument differs. Unincorporated, you'd mirror it contractually: ownership percentages that step up over time, a buyout formula if someone exits early, IP assignment to the venture. But candidly, if the app has real revenue and two owners, the absence of an entity is itself the bigger exposure — liability, taxes, and unsellability. Most founders in your position incorporate and get vesting as part of the same afternoon of paperwork.

AXIS Editorial

Follow-up from maker intake: "My co-founder's lawyer proposed 5-year vesting with a 2-year cliff for me only. Reaction?"

Two separate problems. Longer-than-standard terms are a negotiable preference (5-year schedules exist, especially where founders expect a long build). Asymmetric terms are not a preference — they are a position: your commitment is suspect, theirs isn't. The productive counter is symmetry at whatever length: 'any schedule you'll also be on.' If the answer to that is no, the vesting terms are the least of what you've just learned.