The board sees “execution problems.” The structural diagnosis sees a decision architecture that routes every product decision through the CTO, creating a six-to-eight-week delay on every release. The board sees “culture issues.” The diagnosis sees three teams with conflicting mandates that produce interpersonal friction regardless of who fills the roles. The board sees “the team isn’t aligned.” The diagnosis sees an organisation that never built the translation layer between strategic intent and operational work. One set of observations is descriptions. The other is fixable. That’s the difference between board observation and structural diagnosis — and it’s why the same portfolio company can underperform for eighteen months while the board gives advice that doesn’t land.
Board observation tells you what happened — structural diagnosis tells you why it keeps happening
Board observation operates through curated quarterly updates, compressed narratives, and metrics designed for investor communication. This isn’t deception — it’s structural. The reporting cadence filters out the operational texture where organisational dysfunction lives. A board member sees that sales missed target by 30%. They don’t see the causal chain: product prioritisation is driven by engineering preferences, which means the features customers need don’t ship on time, which means the sales team is selling a product perpetually six months behind market demand. That chain requires mapping the actual decision flow inside the organisation, not reviewing the metrics that result from it. At ICEYE, I watched this gap from the inside — the board data was accurate, the quarterly metrics were real, and the structural cause of the commercial bottleneck was invisible from the board seat because it lived in the decision graph between three functions, not in any one function’s dashboard.
The four structural patterns that generate underperformance
A structural diagnosis examines the organisation as a system producing outcomes. Four patterns account for the majority of portfolio underperformance in climate tech companies, and they compound when they co-occur. Decision bottlenecks — where decisions stall, who holds real authority versus nominal authority, and where the gap between formal and informal decision-making creates latency that the market doesn’t forgive. Role confusion — where competing mandates, unclear ownership, and misaligned incentives make it structurally impossible for individuals to succeed regardless of capability. Strategy-execution gaps — where strategic intent articulated at the board level loses coherence as it moves through the organisation, producing work that’s locally rational but strategically misaligned. And adaptation traps — where accumulated workarounds have become invisible process weight, making the organisation three to four times slower than the work actually requires.
The diagnostic question isn’t “which pattern is present?” — it’s usually all four, interacting. The question is: which one is the root, and which ones are downstream symptoms? A company with a decision bottleneck at the founder will also show strategy-execution gaps (because the translation layer was never built), role confusion (because the senior hires can’t exercise the authority their titles promise), and adaptation traps (because the team has spent years building workarounds for the bottleneck). Fix the bottleneck and three other symptoms diminish. Fix the symptoms without finding the root and you’re reorganising the deck chairs.
The diagnosis produces a mechanical explanation, not an opinion about the team
The diagnostic output is a structural map — not a judgment of who’s good and who isn’t, but a mechanical explanation of how the organisation converts inputs into outputs. Proterra’s August 2023 Chapter 11 wasn’t because the team couldn’t build buses. It was because three businesses — bus OEM, battery supplier, charging network — ran on one balance sheet with three incompatible cadences, and the decision architecture couldn’t reconcile them. AppHarvest’s July 2023 Chapter 11 wasn’t because the greenhouse technology failed. It was because expansion commitments were set before the first facility reached steady-state yield — a decision architecture failure, not an execution failure.
The structural map reveals that underperformance almost always has a specific source, not a general capability deficit. The company doesn’t have an “execution problem.” It has a decision architecture that routes every product decision through the CTO, creating a bottleneck that delays releases by six to eight weeks. That’s specific. That’s fixable. And it’s invisible without a diagnostic methodology designed to find it.
Structural diagnosis starts with the organisation and works up — consulting starts with the strategy and works down
Management consulting typically asks: is the strategy right, and is the organisation aligned to it? Structural diagnosis asks: what is this system actually producing, and why? The difference matters because the strategy is usually fine. Climate tech founders with deep domain expertise rarely have a strategy problem. They understand their market, their technology, and their competitive position. What they have is an organisational design that can’t operationalise the strategy — and no amount of strategy refinement fixes a broken operating system.
Consulting engagements that begin with strategy review and end with a reorganisation recommendation miss the point: the problem isn’t the org chart. It’s the decision mechanics, information flows, and role interactions that exist independent of any org chart. The org chart at Katerra looked reasonable. The vertical-integration ambition — factories, design, construction, technology — was strategically coherent. The organisation that was supposed to execute it was structurally incapable of coordinating across four operating models simultaneously. $2B raised. June 2021 shutdown. The strategy wasn’t wrong. The operating system was.
The cost of running without the diagnosis
The data is consistent: the majority of portfolio failures are attributed to organisational and team issues in the post-mortem. Not technology failure. Not market absence. Organisational dysfunction that was either invisible at due diligence or visible but not assessed. The cost isn’t just the failed investment. It’s the eighteen months between the investment and the acknowledgment that the problem is structural — months spent on board advice, executive coaching, management reshuffles, and COO hires that address symptoms while the structural cause compounds. Eighteen months of a founder hiring and firing senior leaders. Eighteen months of a product roadmap slipping. Eighteen months of the next fundraise pricing in the delays.
A structural diagnosis takes three to four weeks and produces specific, actionable interventions tied to specific mechanisms. Not “improve communication.” Redesign the information architecture so the product team has access to customer data without routing through sales leadership. Not “empower middle management.” Redistribute decision rights so that specific decisions are made by specific people with specific information, at a cadence that matches the business need. The investor’s role shifts from giving advice the founder can’t act on to creating the conditions for structural intervention. For the full assessment methodology, see structural due diligence.
If the last three board conversations have circled “execution” without anyone naming the specific mechanism, the problem is in the decision graph. Run the three-week diagnosis before the next board meeting.