Your cofounder left the company but still has equity, and there was no vesting schedule, so they hold half of everything you build for the next decade. This is the most common founder disaster there is, and it is fixable more often than the forum threads suggest. Not by a clever legal move. By one uncomfortable conversation you have to open correctly.
Here is how that conversation actually goes.
How do I open the conversation without blowing it up?
Open with the ask, the timeline, and an explicit promise not to relitigate why they left. The single most common failure is starting with grievance, because the moment your cofounder is defending their character they stop negotiating over shares.
Say something close to this:
"I want to talk about the cap table. I'm not going to argue about why you left and I'm not saying you did anything wrong. I want us to land on a number we both think is fair, and I want it done in the next two or three weeks so it stops sitting between us. I'll send a written proposal after this call either way."
That's it. No preamble about how hard the last eleven months have been. No spreadsheet of hours worked.
Do this by video or in person, never over text, and never in a shared Slack. And do not open with zero. Founders who anchor at "you should give it all back" almost always end up worse off than founders who anchor at a real number, because zero tells the other person the conversation is a demand and their only move is refusal.
What will they say back, and what do I say to that?
Expect four responses, and you can prepare exact replies for all of them. They will say the split was agreed, that they built the original product, that they took the risk when nobody else would, or that you should talk to their lawyer.
"We agreed on 50/50. I'm not giving my shares away."
"You're not giving anything away. I'm asking you to sell some back and keep the rest. The only questions are how many and at what price."
"I wrote the first version. The idea was mine."
"You did, and that's real, and it's why I'm not proposing zero. What four months of work isn't worth is half of the next six years. I'm trying to price the four months, not erase them."
"I took the risk with you when nobody else would."
"You did, for four months. I'm still in it. If the company sells for something in 2031, you'll be paid for the part you were here for. I want that part written down while we both still like each other."
"I'll have my lawyer look at it."
"Good. Send me their email and I'll send the proposal to both of you. I'd rather they see the numbers than hear about them secondhand."
Never say "you did nothing." Never say "no investor will fund us with you on the cap table" as a threat. Say it as a shared problem instead, and only if it's true: "Here's what came back from the two funds I talked to. This is now blocking the round for both of us." Contempt is what turns a two-week negotiation into an eighteen-month one.
What are the three deals that actually close?
Three structures resolve this in practice: a cash repurchase, a retroactive vesting agreement, or accepting the position and diluting through a priced round. Pick one before the call and know your fallback.
1. Negotiated repurchase. You buy shares back for cash or a note. The number is not the SAFE cap value. If you've raised on a $6M post-money cap, half the company is nominally $3M and you obviously cannot pay that. Real closings land in the low five figures, sometimes a small note paid over 12 to 24 months, sometimes tied to a liquidity event. What gets it signed is a clean cheque now against an illiquid maybe later.
"$25,000, paid $10,000 on signing and the rest over six months, for 40 of your 50 points. You keep 10."
2. Retroactive vesting. Apply the schedule you should have written on day one. Standard four-year, one-year cliff, and credit the months they served. Four months is inside the cliff, which technically means zero, so nobody signs that. The version people sign gives credit plus a bonus for early risk: they keep 5 to 12 percent, the rest is subject to repurchase at the price they paid. Important tax detail: imposing a repurchase right on already-outstanding shares usually calls for a fresh 83(b) election within 30 days, so route it through an accountant before anyone signs.
3. Dilution through a new round. You accept the cap table, raise, refresh the option pool, and their 50 percent becomes 28 percent after two rounds. This is the slow path and it works, but only if investors will fund a company with an absent 50 percent holder, which many will not. And issuing shares mainly to dilute someone is a bad idea legally: under Delaware law, an issuance whose primary purpose is to dilute a stockholder gets reviewed under the entire fairness standard, which is where board decisions go to lose.
Whichever one you pick, get the terms into a document while the agreement is warm, that same week. A structured written agreement covering the share count, the price, the payment schedule, and the release of claims is what makes this real. Servanda is built for exactly this moment, when two people have reached something workable and need it captured in language both will still recognise in a year.
What if they just say no?
If they refuse everything, stop negotiating and go build, because a cofounder holding dead equity has no leverage over your day-to-day and a stalemate costs them more than it costs you over time. Their shares are worth nothing until you create a liquidity event, and you control whether that happens.
Say this and then genuinely mean it:
"Okay. Then we leave it as it is and I'll keep working. My offer stays open for 90 days at these numbers. After that I'd rather spend the money on engineers."
A deadline that you actually honour is more persuasive than any argument about fairness. Many of these deals close in month five, not month one, after the other person has watched the offer sit there and realised no better one is coming.
What you should not do is litigate first. Fraud, breach of fiduciary duty, and misappropriation claims exist, and lawyers will happily quote you $40,000 to start exploring them, but suing your former cofounder over unvested-that-never-was equity is a two-year distraction that scares off every investor who runs diligence on you.
One more thing worth saying out loud on that call, because it costs you nothing and changes the temperature: "Whatever we land on, I'd like you to be someone I can call in five years." Most people, given a route to a fair number and their dignity intact, take it.
