Spiritual Holding Company (Mission Guardian Selection)
Give the mission its own sovereignty by appointing a renewable guardian that outlives any founder.
- Difficulty
- Expert
- Time to result
- ~months to results
- Steps
- 4
- Confidence
- 87%
A governance decision framework for making a company 'mission-controlled' rather than founder- or investor-controlled. You first decide whether the mission guardian is a person or an entity, then choose a structure that can renew its protections over time. Use it when moving past temporary founder control toward durable, mission-locked governance.
Origin
Ries coins 'spiritual holding company' as an omnibus term because advocates of each specific structure infight over which is best. He traces the two-tiered industrial foundation to 1920s Denmark's Nordisk Insulin Laboratorium (today Nova Nordisk), and details Anthropic's Long-Term Benefit Trust, which he advised on in a 'bit part' when Dario Amodei was a first-time founder before the generative-AI boom.
Core principles
- 01A mission needs a guardian because financial gravity won't preserve it by accident.
- 02Founder control is a fine temporary bridge but leaves founders as an un-shruggable Atlas and can't renew itself.
- 03Single-entity rules (Costco) resist attacks but can be chipped away with no way to grow protections back.
- 04The most durable structures have a renewable steward plus a separate enforcer who can sue if trustees deviate.
How to run it
- 1
Decide: person or entity guardian
The first decision point. A person (founder control via dual-class or founders-preferred) works early; an entity (trust or foundation) is more permanent and institutional.
Pro tip Ries argues an entity structure is frankly better than founder control, which carries real downsides and mental-health costs.
Watch out Founder control makes you Atlas holding back the abyss; many founder-controlled leaders end up publicly miserable.
- 2
Choose single-entity vs renewable steward
Either write the rules straight into the structure (Costco's governance fortress) or appoint a stakeholder steward you can renew over time.
Watch out Single-entity fortresses can't regrow protections that get chipped off attack by attack.
- 3
Pick a renewable steward model
Options: nonprofit foundation that owns equity (Nova Nordisk), a non-economic perpetual purpose trust (Patagonia, Anthropic's LTBT), employee ownership trust (John Lewis), or employee voting trust (Alibaba).
Pro tip A perpetual purpose trust can pair trustees with a 'purpose protector' whose job is to sue the trustees if they deviate — checks and balances like a government.
- 4
Start cheap: pledge it into the charter, boot it up later
You don't need to stand up a nonprofit today. Write into the charter that (e.g.) 10% of equity is pledged to a future foundation with 1% of revenue and a board seat — fire and forget, activate later.
Pro tip This is how Anthropic did it: they had the right and intention in their legal documents from inception but didn't implement the LTBT until their Series C.
Watch out Secure the right now; if you wait, you lose the leverage to add it.
In the wild
August and Marie Krogh incorporated the Nordisk Insulin Laboratorium as a for-profit owned and governed by a nonprofit foundation to prevent price-gouging on insulin. Trustees once intervened to stop the for-profit from selling out.
→ The structure has protected scientific integrity for over 100 years; one intervention created over $500B in shareholder value, and such firms are 6x more likely to reach year 50.
Anthropic incorporated as a PBC from inception with the charter right to install a trust, then implemented the LTBT at Series C. It appoints for-profit board directors who are accountable to outside AI-safety trustees with no equity, so they have no financial incentive in growth.
→ The structure let Anthropic turn down a $200M Pentagon contract and refuse to release models — costly moves an investor-controlled board could have blocked by ousting the founder.
Common mistakes
Relying on founder control forever
It's a temporary bridge with a heavy personal toll; without a renewable entity, the mission dies when the founder is decapitated by an activist investor owning as little as 0.5%.
Fighting over the 'one true' structure
Advocates of foundations, PPTs, ESOPs, and no-investor models each think theirs is best; the point is to have a mission guardian at all, not to win the taxonomy war.
Is it for you?
Best for
Founders past the earliest stage who want durable mission protection that survives their own departure.
Not ideal for
Very early SAFE-stage founders for whom a simple PBC filing plus a charter pledge is sufficient for now.
From the transcript
“It has to be somebody or some entity's job to make sure that the thing remains mission locked or mission aligned.”
“I use the omnibus term spiritual holding company to describe this category of things”
“founder control is fine as like a good bridge, a temporary bridge to a more permanent structure”
“All you have to do is write it into the charter the way Anthropic did.”
From the episode
How to build a company that withstands any era
Eric Ries, Lean Startup author