A workshop ends with a list of candidate products and services, each with a first test and an owner. Venture building is what I do with that list next. We pick the few worth real money and test each one with a paying customer. The ones that pass become businesses, inside your company or spun out of it.
I have founded five companies and led one exit, and I co-lead the Startup Studio at Cornell Tech, where about 60 teams a semester take an idea to a validated venture. Venture building runs your candidates through the same method, with your experts and your customers in the room. I cannot tell you in advance which candidates will survive, which is why the gates exist. Most of them will die at one, and that is far cheaper than finding out after the build.
The work runs in three phases, and each one ends with a decision.
We start from the workshop record and from what your business already knows it should build. A candidate gets a team only after a sponsor from your business puts their name on it. Leadership teams have more ideas than capacity, so I push for three to five candidates.
Each candidate gets one customer as a design partner and a small team. The team talks to customers the way Studio teams do, builds a prototype on public or approved data, and comes back with evidence against the five questions every table answered in the workshop, including what delivery costs once a person has reviewed and corrected the output.
A candidate that clears its gates gets a paying pilot and a founding team, which can include founders from the Cornell Tech network. From there it goes to your engineering team or production partner for the real build, or it spins out as a company. When two or three candidates sit on the same underlying system, we build that system once and give each candidate its own product on top.
Every candidate passes four gates. Most of the spend sits after the third, so a wrong candidate dies cheap.
Your business supplies the sponsor, the expert the prototype has to satisfy, and the people on each team. The sponsor owns the decision at each gate.
Your customer supplies the problem and, at gate 3, the money. One design partner per candidate is enough. Two is a committee.
I bring the view of where the technology is going, the prototypes, the customer interviews, and the work of forming a company around a candidate that passes, plus founders from the Cornell Tech network when a team needs them. Your engineering team or production partner takes anything validated into production, so the people who already build for you keep building.
Rules about who can see which customer's data hold the whole way through. Every first test runs on public data, synthetic data, or data a customer has approved for the purpose.
I work for a fee, for a lower fee plus equity in each venture, or for a fee with the option to co-invest. Co-funding is on the table. We settle the structure before the first gate, while every candidate is still a hypothesis and the conversation is easy.
Send a note to alberto@escarlate.com with a line about the business and what you think it should build. If you already have the list from a workshop, send that too. I work from New York and keep a regular agenda in Brazil and Latin America.