Resources
A model that only works for one network is a hobby.
Consilience Ventures ran this shape from 2019 to 2025: scoped work, a shared unit, settlement after acceptance. Execution Capital is that model as infrastructure any ecosystem can run.
Chapter 03 — We tested it before we sold it
Consilience Ventures, 2019 to 2025.
The obvious objection to any new instrument is that it works on a whiteboard. So before Execution Capital existed, the model ran for six years inside a working network.
Consilience Ventures assembled a network of over 300 vetted senior professionals and put them to work inside portfolio companies, paid partly in cash and partly in units of a shared vehicle rather than in equity in the individual company. Companies were backed alongside institutional co-investors. It ran through a full cycle, including two years most early-stage investors would rather forget.
Six years is long enough to find out which parts of an idea survive contact with reality.
What it proved
The parts that held
Senior operators will take a portfolio unit instead of a full cash rate, when the portfolio is real and the selection is credible. Companies will commit economic rights for execution they can actually use. A network of that seniority can be assembled and kept engaged. And the governance templates work.
What it exposed
The parts that did not
Scoping by conversation does not scale — work needs acceptance criteria written before it starts. Unit pricing set case by case creates quiet unfairness between early and late participants. And an instrument without a published policy behind it is a permanent negotiation.
Why we did not keep it to one network.
We could have run one network and kept the mechanism to ourselves. It was working. That is what most people would have done with it.
The first reason is that the problem is not ours — it belongs to every fund with a portfolio it cannot properly support, and every founder holding a round that will not stretch.
The second is that a single network cannot generate the thing that makes the model compound. One network is a proof point. Many networks, each run by a builder with their own thesis and their own standards, is infrastructure.
So we took the mechanism apart, wrote down every rule that had been a judgement call, and rebuilt it as something another VC could run under their own name. The point was never to run the best network. It was to make the mechanism ordinary.
If you were in that network, or want to run the next one.
The company page has the data and the team. VCI™ explains the unit. Or write to us directly.