"Would you build it for equity?" is the question every agency hears and most agencies dread. The dread is rational: badly structured equity deals combine the worst of both worlds — agency economics with venture risk. But structured well, software-for-equity is the most honest engagement model there is. It's the only one where the builder's payday depends entirely on whether the thing works.
Why agencies get this wrong
The standard failure starts with a category error: treating equity as a discount negotiation. The founder asks for a lower price; the agency, wanting the logo or the dream, "takes the difference in equity" — without a scoped estimate, without a defended valuation, without vesting terms. Six months later neither side can say what was bought, what it was worth, or what happens if the roadmap changes. The deal wasn't priced; it was improvised. Improvised deals don't survive contact with a real cap table, and they poison the relationship they were meant to cement.
The second failure is quieter: the agency staffs equity work with whoever is on the bench, because "it isn't revenue." The product gets the B-team, the product fails, and the agency concludes equity deals don't work. They never tried one. An equity engagement staffed below paid-work standards isn't a bet — it's a write-off with extra steps, and the founder is the one who pays for the lesson.
The core discipline: one estimate, two currencies
Every equity deal we do starts identically to a paid engagement: a scoped, milestone-priced estimate at our normal market rate. That number is not a negotiating posture — it's the same figure a paying client would sign. Only then does the currency conversation start: does the founder pay it in cash, in equity, or in a mix?
This ordering matters because it separates two questions that ruin deals when blurred: what is the work worth (an engineering question) and what is the company worth (a market question). The build value is ours to estimate; the valuation is yours to defend — ideally with a recent round or an advisor-backed number. The equity stake is then arithmetic, not arm-wrestling.
Build value ÷ valuation = stake. Everything else in the deal exists to keep that formula honest.
Why hybrid usually wins
A full-equity deal — the entire build for ownership — is the headline structure, but it's the rarer one. Full equity concentrates risk on both sides: the founder takes maximum dilution, and we take maximum exposure to a single outcome. It's reserved for the strongest theses, where our conviction is close to a founder's own.
The hybrid — you pay under price, the discount converts to equity — is what most partners choose, for good reasons. Cash covers our cost base, which means the engagement doesn't compete with paid work for staffing. The founder's dilution stays modest. And crucially, both sides keep skin in the game at a proportion that matches their conviction: the deeper the discount, the bigger our bet.
Run illustrative numbers to see the shape. Say the scoped estimate is 300 and the defended valuation is 3,000 (any currency, same arithmetic). Full equity: a 10% stake, maximum dilution, maximum exposure. A hybrid at half price: 150 in cash, and the 150 discount converts to 5%. The founder keeps five extra points, our costs are covered, and both sides still lose real money if the product doesn't ship. That last property is the point of the whole structure — a deal where either side can shrug off failure is not a partnership, it's a marketing expense.
There's a portfolio truth underneath the structure, and founders deserve to hear it plainly: a build partner can only carry a few equity bets at a time, because each one consumes the scarcest resource — a senior team's quarters. That scarcity is your assurance, not your obstacle. A partner who says yes to every equity pitch is telling you the stake is priced at zero; the one who runs diligence and declines most of them is telling you that when they do commit, they've concluded your company is where those quarters earn the most. Scarcity is what makes the signature mean something.
Vesting against milestones, not time
Standard startup vesting is time-based because employees contribute continuously. A build partner contributes in shippable increments — so our equity vests against delivery milestones instead. Ship M1, vest the corresponding tranche. Miss it, and the unvested equity stays with the founder. It's the same accountability we sell in paid work, applied to ourselves, and it's the clause that most reassures later investors during diligence.
Diligence goes both ways
Founders expect to be diligenced; some are surprised that we expect to be. Before we commit a senior team's quarters to a stake, we examine the same things an investor would. Distribution first: who is waiting for this product, and what evidence exists beyond enthusiasm — signed letters of intent, a waitlist, an audience, a channel the founder already owns? We build products, not audiences, and no engineering excellence compensates for nobody-wants-this.
Then the cap table, which should be readable in one sitting: who owns what, what's promised to whom, and whether an early advisor holding a blocking stake will make every future round an archaeology project. Then unit economics that survive contact with a spreadsheet — not projections of projections. And finally decision speed, the underrated one: a founder who needs six weeks to approve a milestone scope will starve a build partner the same way they'd starve a hire.
What the founder should ask us
The diligence symmetry means you should interrogate the deal from your side, and the good questions are known. Who exactly will build this — names, not job titles — and what happens to staffing when a paying client gets loud? What are the information rights: does the build partner see board materials, and should they? What happens to the stake if the partnership sours — is there a buy-back mechanism, and at what price? Does the partner expect follow-on rights in the next round? A build partner who answers these fluently has done this before; one who improvises will improvise on your cap table too.
The clauses that keep the deal honest
Beyond the stake and the vesting, a handful of clauses decide whether the deal survives contact with reality. IP assignment tracks payment and vesting — the founder owns what has been paid for or vested, cleanly, so a broken partnership doesn't leave the product's ownership in dispute. A rescoping mechanism — startups pivot, and the milestone plan needs a written procedure for changing course: re-estimate, re-price the remaining tranches, both sides sign. Without it, the first pivot turns the equity schedule into an argument. A buy-back option at a formula price — if the relationship ends, the founder can reclaim unvested and even vested equity at terms agreed while everyone still liked each other.
Just as telling are the clauses we don't take. No board seat — we're a build partner, not a governor, and a vendor with board power is a conflict of interest wearing a fleece vest. No anti-dilution protection beyond what common shareholders get — if the company needs a down round, we take the pain with everyone else. And no exclusivity over the company's future engineering: the goal is a product so well-built that they keep choosing us, not a contract that removes the choice.
What we say no to
Ideas without distribution — we build products, not audiences. Valuations that require believing three miracles. Cap tables that need archaeology. And any deal where the founder wouldn't take their own terms if the roles were reversed. A good equity deal should read like a boring contract wrapped around an exciting bet. When it's the other way around, walk.