A venture studio, from economics to operations
The situation
A venture studio starts companies itself instead of backing other people’s. It pays to launch each one, and it carries a shared team that every venture draws on.
It has to know what each launch really costs and what stake it keeps. It also needs to know what the portfolio must return to be worth doing.
The work is underway, and none of it is finished.
Why it is hard
Most new ventures fail, and a studio’s return comes from one or two that do not. The useful question is what the winners must be worth, not what the average venture is worth.
- The studio’s stake shrinks with every round of new money. A large share at founding can be a modest one at the sale.
- The shared team is a real cost that belongs to every venture, but no single venture pays for it.
- Money goes out for years before anything comes back, so when a sale happens matters as much as its size.
- A plan built on the average venture looks safe. It says nothing about whether any one venture can carry the rest.
- The studio’s return is the cash it gets back, not what its ventures are valued at. Paper value does not pay the shared team.
The approach
We are working through the economics first. That covers the cost to launch each venture, the stake kept through each round, and the return needed.
- The cost of one venture: its own launch cost, plus its share of the shared team.
- The stake: the studio’s share at founding, less what each round sells to new investors.
- The target: a multiple of everything the studio spends, not of what it spends on any one venture.
- The winners: the sale each successful venture must reach to hit the target, at different success rates.
It also covers the shared layer: the team and services every venture draws on. That means what they cost, and how the cost is charged.
That part comes after the economics.
Every figure sits in one model. A change to the launch cost or a round’s size flows through to the sale each winner needs.
A worked example
This is a synthetic studio, not the client’s. It launches six ventures a year, each costing 0.6 million, with a shared team costing 1.2 million a year.
| Step | How | Figure |
|---|---|---|
| Cost to launch | the venture itself | 0.6 million |
| Share of the shared team | 1.2 million a year across 6 ventures | 0.2 million |
| Cost of one venture | launch and team | 0.8 million |
| Stake at founding | the studio starts it | 70 percent |
| After the first round | 20 percent sold to new investors | 56 percent |
| After the second round | 20 percent sold | 45 percent |
| After the third round | 15 percent sold | 38 percent |
Each venture costs 0.8 million once its share of the team is counted. A 70 percent stake at founding falls to 38 percent after three rounds.
A year of launches costs 4.8 million. To return three times that, the winners must bring back 14.4 million to the studio.
If one venture in six succeeds, each winner must sell for about 38 million. At one in three, about 19 million would do.
At one in ten, each winner needs about 63 million. The number of winners matters more than any other input.
The other input that moves it is the stake. Every point the studio keeps through the rounds lowers the sale each winner needs.
Where it breaks
- It treats every winner alike. In practice, one very large sale usually carries the whole portfolio.
- There is no timing yet. Three times the cost over ten years is a much weaker result than over five.
- Round sizes are assumed. A studio that puts money into later rounds keeps more of each winner, at more cost.
- The shared team’s cost is flat. It will grow as the number of ventures grows.
If you are starting a studio, work out the sale each winner needs before choosing how many to launch.