Venture studios work like a production line, not a portfolio. The studio picks the problem, kills most of the ideas it generates against that problem, funds the survivor out of its own balance sheet, hires the founder into it, builds the first version with its own staff, and only then lets the company become a separate entity with its own cap table. Money goes out for months or years before a company exists to receive it.
That sequence is the whole model, and it is the part most explainers skip. This piece walks the process stage by stage, with the costs and decision points at each one, then covers the version that runs on a different clock entirely: building a company out of science rather than software. If you need the definition and the economics first, we cover what a venture studio is separately.
Step 1: Where the idea comes from
Ideas enter a studio through one of three doors, and the door changes everything downstream.
Studio-led. The team generates ideas internally from trend analysis, operator interviews and written theses about what should exist. The founder is recruited into the idea after the studio has already decided it is worth building.
Founder-led. A domain expert arrives with an insight and the studio decides whether to co-build around it. Founders Factory accepts both modes. Here the studio behaves like a sophisticated co-founder rather than an originator.
Corporate-partner. A large company funds the studio to build ventures adjacent to its core business. Forward 31 was built by Porsche Digital on this basis.
Ask which door you are walking through before anything else, because it sets the defensible equity range. A studio that wrote the thesis, ran the research and recruited you into it has a claim on the company that a studio reacting to your idea does not.
Step 2: The funnel that kills most ideas
The filter is the product. Studios do not outperform because they pick better ideas; they outperform because they kill bad ones early and cheaply, before anyone is emotionally or financially committed.
Operator accounts of that filter converge on five gates:
- Thesis screen. Does the idea fit the studio's sector and its existing buyer access? Most ideas die here.
- Desk validation. Market sizing, competitive density, regulatory landscape, technical feasibility. Unglamorous spreadsheet work that kills ideas which looked good in a workshop.
- Customer validation. Twenty to thirty buyer conversations run to falsify the premise, not to confirm it.
- Build and test. A prototype, a landing page, sometimes a manual version of the service, answering one narrow question about use or willingness to pay.
- Founder match. Only survivors of the first four gates get a founding CEO.
The methods have names. High Alpha pioneered a Sprint Week in which the studio shuts down operations for five days to validate or kill a concept, and teams run smoke tests or painted door tests, landing pages for products that do not exist, to measure click-through and acquisition cost before writing code. Engineers move across several projects at once rather than sitting idle between builds. Pioneer Square Labs kills 90 percent of its ideas before they ever take external funding.
The kill criteria matter more than the process diagram. Set in advance, they stop sunk cost from driving the decision. Set afterwards, they are opinion.
Step 3: How the founder gets matched
Founders enter through an entrepreneur in residence program, paid three to six months to explore ideas alongside the team; through direct recruiting into a pre-validated concept, which looks like a senior executive search; or through founder-led intake, where the studio folds an existing founder in with resources behind them. Some studios claim to accept the third route and in practice do not, so ask specifically.
By the time a founder is matched, one operator account puts the studio's spend at 50,000 to 250,000 dollars already sunk into killing or de-risking the concept. That spend on dead ideas is the real cost of the model, and it is why the studio arrives at the table with a number in mind.
The science version is slower and more generous on time. Deep Science Ventures gives founders up to 18 months to form a company, funded for the full duration, working on opportunities the firm has already scoped, with the option of forming more than one company across that period. That is not a program with a demo day. It is employment with an equity outcome attached.
Step 4: The build, and who is actually doing the work
This is the stage founders picture as product development and which is mostly not. A studio operator describes the phase before incorporation, which his firm calls Shape, as one where at least seven different core team members contribute to the business every week, working through dozens of tracked goals covering naming, customer journeys, first product and budgets. Two of those goals decide everything: partnering with a technical co-founder, and signing the first ten launching customers.
The honest shared-services picture is narrower than the pitch deck, and operator accounts are consistent about where the line falls. Legal templates, standardized incorporation documents, recruiting pipelines, bookkeeping, payroll and cap table management centralize cleanly across a portfolio. Design capacity rotates. Engineering does not share well, because code and architecture are company-specific and a shared engineering team slows every company down. Product strategy and customer relationships never share; those are the founder's job and the company's asset.
Healthy studios are explicit about which side of that line each function falls on. Get the list in writing before the equity is agreed.
Step 5: Incorporation and spin-out are two different events
The most common misreading of the model is that spinning out means registering the company. It does not. For a disciplined studio, incorporation happens when the odds justify a legal entity, and the same operator quoted above defines spin-out by a different trigger: the company raising outside funding. Between those two events the studio is still paying for most of the work, and while the first round is being raised it hires the three to five people who will make the company a standalone team on day one.
The cap table gets built across those same stages. At separation, studio-born companies typically land in four bands: the studio at 20 to 60 percent, the founding team at 20 to 50 percent, an employee option pool at 10 to 15 percent, and new investors at 5 to 30 percent depending on the round. Founder equity almost always vests over four years with a one year cliff, and the studio's stake dilutes as the company raises. The number itself deserves its own scrutiny, which is why we break down how much equity venture studios take separately.
What it costs the studio to get one company out
Numbers make the model legible. The same operator puts the opportunity cost of getting a single company to incorporation at 50,000 to 200,000 dollars for a mature studio, and the amount the studio finances after incorporation, before external funding lands, at 300,000 to 500,000 dollars.
At the studio level, a conventional outfit with five to ten team members and three to five active ventures spends roughly 2 to 5 million dollars a year on salaries, infrastructure and venture costs before any venture produces meaningful revenue. It gets paid years later, if at all, which is the structural weakness of the model and the reason studios need either a fund, a corporate sponsor or a profitable anchor business behind them.
Read those figures as a test. A studio spending this much per company has earned a large stake. A studio offering introductions and a template pack has not. The revenue side of that equation is covered in how venture studios make money.
How the sequence changes when the venture is science
Almost every studio playbook assumes software. A corporate studio reports three to twelve months from initial exploration to venture launch. You cannot run a gasifier, a biomanufacturing process or a carbon removal method on that clock, and you cannot smoke test a physical plant with a landing page.
So the deep-tech builder reorders the process. Deep Science Ventures starts from a desired outcome, narrows to a specific opportunity area, recruits a scientist founder into its own team, deconstructs the area from first principles against the literature, builds the team around the optimal technical route, and tests the riskiest aspects pre-incorporation using proven knowledge and IP before the venture is created and invested in. Run at scale, that method produced 35 companies in five years across direct air capture, reforestation, antibody production and compute, with a target of 200 more.
Planning runs backwards rather than forwards. For most hardware sectors you target first commercial deployment around Series B and work back from there: a large pilot at Series A, a demonstrator that removes integration risk before it, and science risk taken off the table at pre-seed. Four risk buckets, commercial, technical, team, and fundraising and operations, change state at each of those stages, and because the cost of pivoting a hardware company is brutal, the milestones for the next five to seven years are set before the build starts.
The build phase looks different too. A scientific venture studio runs four phases per venture and considers hundreds of opportunities per cycle to select around ten, with the builder acting as a rounded interim C-suite that hands over a funded and staffed company. Where the science comes out of a university, the spin-out involves the tech transfer office and the IP assignment, and the builder needs six months to a year from spin-out to a credible pre-seed proposition, with a data room at the depth investors normally expect at Series A.
Then there is the part no software studio has to solve: manufacturing, shipping, installing and servicing hardware in the country where it runs, filing patents before publication destroys novelty, and sequencing grant money against equity so dilution stays sane across a decade. That is the model EX EPIC operates, and the portfolio numbers are what this version of the process produces: 200 or more waste-to-energy units deployed across 11 countries, 24 patent families filed with roughly 100 more validated in the pipeline, 250 or more operators trained into the portfolio, on a capital track record above 160 million euros across four continents. We go further into that distinction in a deep-tech venture builder.
What to ask before you build with a studio
Six questions, drawn from the stages above, that separate an operator from a brand:
- Which door did this idea come through, and who did the validation work?
- What are the kill criteria, who applies them, and at which gate?
- How many people are on the bench for this build, and for how long?
- Does spin-out mean incorporation or does it mean an external round?
- What does the cap table look like on separation day, including the option pool?
- What is your role after the round closes, and is it a board seat or an operating one?
J.P. Morgan names the founder-side costs of the model plainly: equity dilution, reduced founder influence over decisions, and the risk that the studio's priorities diverge from the company's. Those are real, and they are priced fairly only when the studio is genuinely doing the work described above. If you are still choosing between models, we compare studio, accelerator and incubator side by side.
FAQ
Do venture studio founders get paid a salary? Usually, yes. Entrepreneurs in residence are paid for a three to six month exploration period before any company exists, and in the genuine co-founder band the studio pays the founding CEO a salary alongside taking equity. Deep-tech creators go further, funding founders for as long as 18 months while the company is being formed. Confirm who pays the salary after incorporation, because that often shifts to the venture.
Who owns the IP before the company is incorporated? Not the venture, because it does not exist yet. The studio holds or licenses what it uses to de-risk the concept, and where the science comes from a university the rights sit with the institution until a transfer is negotiated through its tech transfer office. The assignment lands on the new entity at spin-out, and getting that sequence wrong is one of the more expensive mistakes in deep tech.
What happens to the studio after the company spins out? The core team moves from operating the company to supporting it. Handovers are done, roles the venture now needs full time have already been hired, and a studio partner typically stays on the board alongside the CEO and CTO. Builders that stay operational after the first substantial round are usually signalling that the company was not ready to separate.
How many ideas does a studio look at for each company it launches? Far more than it builds. One studio kills 90 percent of its ideas before external funding, and a scientific venture studio may screen hundreds of opportunities in a cycle to select roughly ten for a partner. If a studio cannot tell you its kill rate, it probably does not have a funnel.