Overgrown

Early-stage startups operate under brutal constraints. Never enough time. Never enough people. Never enough money. You’re default dead, racing against the clock to earn your right to continue.

Then you survive, and you grow. One product team becomes four. An initiative gets staffed with five designers producing “activity” without “motion.” Revenue keeps climbing and the runway that once felt desperate now stretches for years. Project timelines stretch right along with it. The existential pressure that drove you here fades, and with it, something else fades too: urgency.

“Necessity is the mother of invention,” Plato wrote. What happens when necessity disappears? You have to manufacture it by creating a greenhouse for incubating your innovation initiatives.

From the Lab to the Greenhouse

The typical corporate response to the urgency problem is an innovation lab. I’ve seen this model up close. It creates a permanent class of innovators who build prototypes and hand them off to production teams. Context evaporates, the org splits into haves and have-nots, and without real constraints the lab becomes a fiefdom where activity masquerades as progress.

The greenhouse offers a different model. Most people think greenhouses are about protection. But real greenhouses don’t just protect. They manufacture growing seasons. As Orson Welles put it, “The enemy of art is the absence of limitations.” Greenhouses impose real constraints with real consequences, and the isolation is temporary. Teams earn their right to continue or get composted.

Controlling the Climate with Manufactured Constraints

An internal board of directors can act as venture capital, placing bets on teams with explicit constraints:

Limited time. Think of timeboxing as creating an artificial season. Plants don’t have unlimited time to grow. Winter is coming. That pressure drives them to flower and fruit.

Your greenhouse teams need the same pressure. A clear timeline with clear success criteria. Miss your targets and you’re “off track.” Repeatedly off track without a credible path forward? The initiative gets composted. This is the model adopted by many startup incubators, and the pressure is the point.

Limited resources. Hard caps on team size, budget, and headcount. A squad of five to seven engineers costs roughly a million dollars per quarter. Make that number visible. When you can’t throw money at problems, you solve them differently, or you learn the initiative isn’t worth the investment.

Incubation Creates Insulation

Organizations will try to kill new ideas. Not maliciously. Just people doing their jobs. Architecture review wants consistency. Security needs threat models. Legal needs compliance sign-off. Product management wants roadmap alignment. Each request is reasonable in isolation. Together, they suffocate anything new.

Ed Catmull nailed this in Creativity, Inc.: “Part of our job is to protect the new from people who don’t understand that, in order for greatness to emerge, there must be phases of not-so-greatness.”

For innovation teams, the greenhouse walls provide insulation from these organizational processes. Architecture review. Sprint ceremonies. SLAs. Code coverage requirements. Documentation standards. These processes exist for good reasons in the main organization. Inside the greenhouse, they’re waste. The walls keep them out and the team decides what is “good enough” for their customers.

Build the right thing first. Build it right second.

Like a bio-dome, contact with the outside world should be minimal. Create a small set of explicit touchpoints (weekly demos, monthly board reviews) and then get out of the way. The constraints do the managing from there.

Before the walls go up, align on the minimum bar for viability inside the greenhouse. Security is a common one. Even a prototype needs guardrails that protect the broader ecosystem. Agree on the floor early so the team can move fast on everything else.

The Gardeners

Constraints only work with the right people inside the greenhouse.

You need a small, focused group of generalists willing to do whatever it takes. Backend, frontend, user research, sales calls. Everyone does everything, with minimal support from specialists outside the greenhouse. This stretching may be uncomfortable for some. That discomfort is a feature, and most closely resembles the DNA of early-startup talent.

One master gardener. Someone who owns the whole thing, makes fast decisions without waiting for consensus, and won’t shut up about it. Not a committee. One person.

The team stays with the plant not the greenhouse. They go in together and come out together, whether that means transplanting into the main organization or composting and returning to other work. There is no permanent class of greenhouse innovators. That’s the innovation lab trap, and the greenhouse is temporary by design.

The Greenhouse Lifecycle

Every greenhouse project ends one of two ways: transplanting or composting.

When an experiment proves out, transplant it. The walls come down and the team begins hardening the seedling for the outside world. (Gardeners call this “hardening off” - acclimating plants to outdoor conditions before moving them outside.) The same people who built it are the ones who scale it. No handoff, no knowledge lost.

When an experiment doesn’t get traction, compost it. Nine out of ten startups fail, and internal innovation initiatives are no different. But the speed to learning is what matters. When a project gets composted, the learnings don’t die with it. In nature, nothing is wasted. Dead plants decompose into nutrients for the next generation. Deliberate “composting” means extracting what you learned about the market, the technology, the approach, and distributing that knowledge back into the organization. Active fertilization, not a post-mortem that gets filed and forgotten.

Ninety percent failure rate sounds bad until you realize the alternative: slow-moving initiatives that consume resources for years without ever producing a clear signal. Fast failure with fast learning beats slow failure with no learning.

Time to Plant

Build a greenhouse. The walls protect the new from the corporate immune system. The constraints create fruitful growing seasons. Together, they set up the conditions for innovation to thrive.