Teardown
Teardown: Y Combinator’s Standard Deal
This one breaks a structural assumption the checklist has carried since Linear: that the customer pays cash for something and Price is a dollar figure on the page. Y Combinator’s standard deal inverts that. YC gives a founder five hundred thousand dollars in cash. The founder gives up equity in return. Price, in the ordinary sense, runs backward.
The offer as written. Y Combinator invests five hundred thousand dollars in every company accepted into its batch program. The money comes in two pieces. A hundred twenty-five thousand dollars buys a fixed seven percent of the company through a post-money SAFE, an ownership percentage set at the moment of investment regardless of what the company is later worth. The remaining three hundred seventy-five thousand dollars comes through an uncapped SAFE carrying a most-favored-nation clause, meaning its eventual conversion price tracks whatever terms the company’s next priced round sets. Admission is by application and interview, not by payment, and acceptance rates are low.
Name the Promise, and size it. The Promise is not a feature or a function, it’s a trajectory: capital, mentorship, a peer cohort, and a credibility signal strong enough to change how later investors and the market treat the company for years afterward. This is the largest Promise size seen across five teardowns. It is not bounded to one function like Stripe, not bounded to one team’s workflow like Linear. It claims to change the company’s whole future path.
Identify where Credibility is sourced. Almost entirely from track record made visible after the fact rather than claims made before it. Ten IPOs from a single class year, cited directly in Y Combinator’s own announcement of the current deal terms, is the kind of evidence that only exists because the mechanism has already run many times. A promise this large, made by a newer or unproven program, would have almost nothing to stand on. Here, the credibility isn’t argued for on the page, it’s assumed, because the applicant already knows the track record before they apply.
Measure Toll on both dimensions. Friction is severe and front-loaded: a competitive application, an interview, and for those accepted, relocation and a fixed batch schedule that reorganizes the founder’s life for months. Complexity, by contrast, is unusually low for an instrument this consequential. Two SAFEs, one fixed percentage and one uncapped with one added clause, is a simple structure by startup financing standards. This is the inverse of HubSpot, high friction paired with low structural complexity, rather than HubSpot’s low friction paired with high structural complexity.
Check whether Toll is deliberate. Entirely deliberate, and functioning as the actual product. The application and interview process is not overhead standing between the founder and the offer, it is the exclusivity mechanism itself, discussed further below. Making the process demanding is what makes acceptance mean something.
Name the Best Alternative. For most applicants, it is not a competing accelerator. It is either bootstrapping without outside capital, or raising an equivalent amount from angel investors on negotiated terms. Compared against that alternative, the SAFE terms are only part of the decision. The rest is whether the founder believes the credibility signal and network are worth more than the equity given up, which is a judgment about the Promise from step 1, not about the price itself.
Run the complement test. Doesn’t apply in its usual form, since there’s no bundle of separate components being evaluated for whether the founder would assemble them separately. The closer question is whether capital and mentorship function as genuine complements here, and they do: capital alone is available elsewhere on possibly better terms, and mentorship alone has little value without the capital and cohort structure around it. The two pieces are worth more combined than either would be purchased separately.
Locate the Value Margin engine. This is where the buyer and seller roles genuinely invert. YC is paying cash out and receiving equity, the reverse of every prior teardown’s direction. YC’s own margin depends on the aggregate future value of the equity stakes across an entire batch, most of which will be worth little or nothing, funded by a small number of large outcomes. This is a portfolio-level Value Margin calculation, not a per-transaction one, and it explains why the fixed seven percent piece exists: it guarantees YC a minimum stake in every company regardless of how that company’s later fundraising goes, which is the equivalent of a cost floor in a business where almost every individual unit could otherwise return nothing.
Check Price visibility. Fully visible and precisely specified, seven percent for a hundred twenty-five thousand dollars, an uncapped SAFE for the rest. There’s no ambiguity in the terms themselves, only in what the equity will eventually be worth, which is a function of the company’s future performance rather than of anything hidden in the offer.
Look for filtering built into the price structure. This is where Y Combinator’s design diverges most from every other teardown. The filtering isn’t in the price at all, since the price is the same fixed offer for every accepted company. All the filtering happens before price ever applies, in the application and interview process itself. That inverts HubSpot’s model, where the price structure does the filtering after the offer is already visible. Here the offer only becomes visible to those who already cleared the filter.
What this explains that a term-sheet summary wouldn’t. A term-sheet summary would state the SAFE mechanics and stop. The framework explains why the terms are simple while the process to reach them is difficult: a Promise this large needs an exclusivity mechanism to be credible at all, and here that mechanism sits entirely upstream of price rather than inside it.
What this run adds to the checklist. Two structural additions, both bigger than anything the first four teardowns required. First, Price can run in either direction. The framework has assumed a customer pays a seller. Here the seller pays the customer, and equity moves the other way. Step 7’s Value Margin question needs to ask which party is actually bearing the payment risk before assuming the usual direction. Second, filtering can happen entirely before Price rather than through Price. Step 9 assumed pricing structure itself does the filtering work. Y Combinator’s filtering is complete before any financial terms are presented, through admission rather than through tiering or a cliff. The checklist needs to check for a pre-price exclusivity gate as a distinct case from a price-structure filter, since they produce the same effect through opposite mechanisms.
Where This Sits in Offer Physics
Concepts referenced