A deal both sides want and neither quite trusts
Joe is sixty-two. He’s run Joe’s Auto Body for thirty years. His wife has been ready for him to stop for five. Marcus has been his head mechanic for twelve, and he’s been running the shop in all but name for the last four. Marcus wants to buy it. Joe wants to sell it to him.
They’re not fighting. They’re just stuck on the math.
Joe is asking $350,000 for the business — goodwill, equipment, the customer list, the name on the sign. The building adds another $220,000. Marcus can put together about $220,000 from savings, family, and a bank loan against his house. The gap is more than double what he has.
Seller financing is the obvious answer. They both know it. But then the layers start coming off. Joe’s son is a lawyer, and he keeps pushing for a five-year non-compete across the whole state — in case Joe slides back into the shop a year later and takes the fleet accounts with him. Marcus isn’t going to sign that; if the shop collapses he needs to be able to feed his family doing the one thing he knows how to do. The four commercial fleet accounts are 40% of the revenue, and honestly, they’re friends of Joe’s more than they are customers of the shop. And nobody has figured out what to do with the name.
Neither of them has written any of this down. They’ve been circling each other in the office for months.
What Mediator surfaced
Joe and Marcus each worked through their side privately with their own AI assistant; neither saw what the other produced. What came back wasn’t on either man’s list. It was a version they could both sign:
A non-compete that unwinds if the deal itself fails.
That’s the piece that unlocks everything else. Joe’s three-year, 25-mile non-compete — tight enough for his son, narrow enough for Marcus — terminates automatically if Marcus misses six consecutive payments and Joe has to repossess the business. If the deal collapses, Joe doesn’t get to use the non-compete as a second way to extract value; Marcus isn’t locked out of his own trade while watching the business he sank his savings into fold. A reciprocal two-year, 15-mile clause applies to Marcus if he walks away voluntarily — protecting Joe’s ability to resell or restart the shop.
$280,000 for the business, $80,000 at closing and the rest over eight years at 5% — a monthly number Marcus can actually make. If revenue drops more than 15% in any year, he can defer three months of payments to the end of the term. And if two of the four fleet accounts walk in the first eighteen months and revenue drops 20% with them, $30,000 of his remaining principal is forgiven — provided Marcus made real efforts to keep them. Joe is acknowledging out loud that part of the purchase price reflects relationships Marcus can’t fully inherit.
The building is a separate deal. Joe keeps it, leases it to Marcus at $2,200/month, and Marcus has an option to buy it outright at any point over the next five years, with half the rent he’s paid crediting against the price. Joe gets income if Marcus doesn’t buy; Marcus gets a path to ownership if he does.
And then the part Marcus had been most afraid to ask for out loud:
Twelve months, two days a week, $400/day, focused on fleet-account handoffs. At the end of those twelve months, Joe doesn’t come back to the shop unless Marcus asks him to, for a defined reason, at a consulting rate.
Marcus had privately named the fear that Joe would “keep coming in, informally, to help, and I’ll never fully be in charge.” The agreement writes down the terms under which that doesn’t happen.
Joe was always going to drop by the shop after the sale. The agreement just names the terms — which is the part that lets him actually retire, by giving him a defined role to stop showing up to rather than a vague “I’ll still be around” that turns into five years of hovering.
The agreement above was generated by the live Mediator.ai engine from Joe's and Marcus's private statements. See them here: Joe's · Marcus's