Carbon removal companies are trying to operate as a research lab, a hardware manufacturer, and a project developer inside one building. Three different businesses with three different cadences and three incompatible ideas of what “done” means. The founder sits on top of the collision and calls it scaling.
The science team thinks “done” means publishable and reproducible. The manufacturing team thinks “done” means the BOM is stable and the yields are predictable. The project development team thinks “done” means the offtake is signed, the permits are cleared, and construction starts on the date the financial model assumed. These aren’t disagreements about priorities. They’re three different operating systems running in the same organisation, each optimising for its own definition of success, none of them aware that their definitions are mutually incompatible.
That’s the collision. Everything else — founder transition friction, scaling breakpoints, hire mismatches — is a symptom of it.
The cadence mismatch is the mechanism
Research iterates on a timescale of weeks to months. A variant of the reaction gets tested, the data comes back, the methodology gets refined. Failure is acceptable — it’s data.
Manufacturing iterates on a timescale of quarters. The equipment gets dialled in, the yields move from 70% to 85%, the cost per tonne falls by 15%. Failure is tolerable but expensive, and “frozen design” becomes a load-bearing concept.
Project development iterates on a timescale of years. A site gets identified, a permit gets applied for, a power-purchase agreement gets negotiated, an offtake gets signed, construction begins. Failure is catastrophic: a failed permit or a collapsed offtake doesn’t delay the project, it can kill the company’s next funding round.
Most carbon removal companies are organised around the founder’s native cadence. Scientist-founder companies run on research time — everything iterates, everything is provisional, nothing is frozen long enough for a manufacturing team to hit yield targets or a project team to close permits on a stable design. Project-developer-founded companies run on years — everything is planned, deviation is treated as crisis, and the research team operates under an implicit freeze that kills the iteration speed the science actually needs. Neither cadence accommodates the other two.
This is a specific instance of decision architecture failure. The company needs three different decision cadences running in parallel, with clean authority boundaries between them. What it usually has is one cadence routing every decision through the founder, which means research decisions are waiting on project meetings and project decisions are waiting on lab results — and the organisation’s throughput is capped at the slowest of the three cadences applied to all of them.
The offtake is a contract against a capability that doesn’t yet exist
The structural weight that pulls everything tight is the offtake. A removal company sells future capacity — tonnes of CO₂ that will be captured years from now, at costs the company hasn’t yet achieved, using processes that are still being optimised. Every offtake agreement is a contract against a capability that doesn’t yet exist. The organisation is simultaneously promising delivery, designing the process that will deliver, and building the facility where the delivery happens.
This is what makes the three-business problem non-negotiable. A lab can operate without a manufacturing function for years. A manufacturer can operate without R&D for years. A project developer can operate without a lab for years. A carbon removal company can’t operate without all three simultaneously for any sustained period, because the offtake is the thing that creates revenue, and the offtake depends on all three functions executing against each other’s timelines. A research breakthrough that improves yields is worthless if it arrives after the facility design has been frozen for permitting. A permitting win is worthless if the cost structure still assumes a process variant the science team has since rejected.
The companies that navigate this successfully treat the organisational transformation with the same seriousness as the technical one. They build an explicit decision architecture between the three functions before the collisions become crises. They hire operational leadership with industrial project development experience — not because the founder is bad, but because the cadence mismatch requires someone whose native operating rhythm is years, not weeks. They accept that the founder leadership transition isn’t optional; it’s the structural precondition for scaling at all.
The nominal-authority COO pattern is the most common failure mode
The pattern is recognisable from inside: the founder has hired a COO. The COO has a title and no real decision rights. The research team still routes every technical question through the founder. The project team has been told to own its cadence but every significant decision requires founder approval because the founder doesn’t yet trust that the project team understands the science. The COO’s authority is nominal and the organisation keeps operating on the founder’s cadence. Six months later, the COO leaves. The founder concludes the hire was wrong.
The hire wasn’t wrong. The handoff wasn’t a handoff. The founder transferred the title without transferring the authority, which meant the company went from “one cadence routing through the founder” to “one cadence routing through the founder with extra meetings.” This is a specific variant of scaling breakpoints — the company hit the structural limit of a single-cadence operating model and responded by adding a person instead of adding an operating model.
The companies that break this pattern do one of two things. Either the founder explicitly hands off project development with real decision authority and accepts that some decisions will be made differently than they would have made them. Or the founder moves into a CSO role and brings in an operating CEO whose native cadence is years and whose authority is actual. Neither is comfortable. Both are easier than watching the offtake book collapse because three incompatible cadences were never reconciled.
What investors should be asking about carbon removal companies
For investors looking at carbon removal portfolio companies, the question isn’t “is the technology viable.” Most of them are. The question is: does this organisation have a decision architecture that handles three incompatible cadences, or is everything still routing through the founder’s research cadence? The answer is visible in how the company talks about its COO, how it describes the handoff between lab and facility, and whether the founder can name a single significant decision the project team made last quarter without founder involvement. That’s the diagnostic signal that predicts whether the company converts offtake into delivery.
If your carbon removal company has signed offtakes, hired a COO, and still can’t freeze a design for the first commercial facility, the bottleneck isn’t the process. Three operating cadences are still negotiating through a single founder, and the handoff that needs to happen is not a delegation — it’s a cadence transfer. Map the cadence structure before the next offtake review.