Your due diligence reads the pitch deck, the cap table, and the CTO’s LinkedIn. It checks the technology. It validates the market. It interviews the team. Then it commits capital — without ever evaluating the one layer that determines whether the capital converts into the outcomes your thesis describes: the decision graph.
The decision graph is how the organisation actually makes choices. Not the org chart, which describes roles. The actual structure: which decisions stall at which seams, who holds informal authority that doesn’t match their title, where information gets filtered before it reaches the people who need it, and where the formal and informal systems contradict each other in ways that create exactly the execution gaps you’ll later attribute to “people issues.”
This is the layer where your capital either becomes the outcomes your thesis assumed, or quietly doesn’t. It’s not a team question. It’s a structural assessment with specific, observable markers. And it’s absent from standard due diligence not because it’s hard to do, but because the DD playbook was built for software companies whose operating model is simple enough that the decision graph rarely matters. Climate tech is not that kind of company.
The software DD playbook stops working when the operating model is hybrid
Software DD is compressed because software operating models are compressed. Product, engineering, sales, customer success. Four functions, one reporting line each, a single cadence (the sprint), and a single commercial motion (SaaS). You can evaluate most of the machine from the founder’s calendar and the sales dashboard. If the team slide looks strong, the team slide probably is strong, because the operating model doesn’t generate many structural contradictions between the functions.
Climate tech companies run four operating models simultaneously inside one org. Hardware development on a years cadence. Software development on a weeks cadence. Government contracting on a procurement cycle that doesn’t care about either. Commercial sales on whatever cadence the customer dictates, which is usually none of the above. Every one of these has its own decision authority, its own information flows, and its own definition of “shipped” and “done.”
Standard DD sees four teams on an org chart. It doesn’t see that the four teams require four different decision cadences, that the founder is usually the only person with authority that crosses all four, that therefore every cross-function decision routes through the founder, and that therefore the company’s throughput is structurally capped at the founder’s weekly decision bandwidth — regardless of how good the team is, how large the TAM is, or how compelling the technology is.
I know this shape from the inside. I built €20M ARR in climate adaptation products at ICEYE. The technology worked. The market existed. The team was excellent. The bottleneck was always the same: decisions that required coordination across three of the four operating models had no clean authority structure, and everything routed through the founder because the founder was the only person whose mandate crossed all four. The throughput of a 200-person company was structurally capped at one person’s bandwidth. That cap isn’t visible in a team slide, and no amount of hiring fixes it — because the next senior hire inherits a mandate that still doesn’t cross the four models.
Four markers predict whether a climate tech company can scale
Organisational DD isn’t a team assessment dressed up as a spreadsheet. It’s a structural evaluation with four specific, observable markers:
Decision routing. Ask the founder to describe how a specific product decision got made last quarter. Who proposed it, who approved it, who executed it, how long each step took. Then ask the same question of two other senior leaders independently. The gap between the three answers tells you whether the decision graph is legible or whether it’s improvised daily. A legible graph produces three compatible accounts. An illegible one produces three people telling three different stories and each one believing theirs is the definitive version.
Authority-title alignment. Look for the gap between who holds the actual decision rights on hiring, firing, product prioritisation, and capital allocation — and whose titles suggest they should. A company with a CFO who approves expenses and a founder who approves everything else is a company where the CFO title is decorative. The founder-CEO assessment is where this gap becomes visible at the top of the org; organisational DD makes it visible across the senior layer.
Information asymmetry at the seams. Climate tech companies fail most often at the seams between hardware and software, between R&D and commercial, or between domestic and international markets. Look at what information crosses those seams, at what cadence, and in what format. If the only person who has the full picture of any one seam is the founder, the seam isn’t structural — it’s tribal. Tribal seams work until the person leaves, and then the seam becomes a cliff.
Scaling readiness. If the company needs to go from 30 to 80 people to hit its milestones, does the organisational infrastructure for that transition exist? Or is the plan “hire great people” and assume the structure will emerge? The absence of an explicit structural plan is the plan — and it’s a plan to hit scaling breakpoints at exactly the moment the thesis needs the company not to.
What the assessment reveals that a reference check doesn’t
Reference checks ask: are these good people? The organisational assessment asks: does this system let good people succeed? The distinction matters because talented individuals routinely fail inside broken structures. A strong VP of Sales will underperform in a company where product prioritisation is driven by engineering and sales has no input. An excellent CTO will look indecisive in a decision architecture where the founder overrides every technical call. These are structural failures that look like people problems from the outside, which is why the standard diagnostic — fire the person, hire a replacement — doesn’t work. The replacement will fail for the same structural reasons, and the cost of learning this is another eighteen months.
An assessment that maps the decision graph surfaces the structural cause of the underperformance that your board advice will spend the next eighteen months trying to address. By that point, the structural cause has compounded. The founder has hired and fired two senior leaders. The product roadmap has slipped twice. The next fundraise has priced in the delays. The cost of the assessment you didn’t do is the difference between a company that could have scaled and one that’s now raising a down round and calling it “a strategic round.”
The assessment is a two-week sprint, not a consulting engagement
The organisational DD runs in two weeks, in parallel with the financial and technical assessments. It doesn’t produce a deck. It produces a structural map of the decision graph, a risk profile with named mechanisms, and a set of conditional interventions — “if this structural risk concerns the investment committee, here are the three specific conditions that would mitigate it post-close.”
The output informs both the investment decision and the post-investment plan. If the decision graph is clean, the assessment gives you confidence to move faster. If it isn’t, the assessment gives you either specific conditions for the term sheet or a reason to decline. For the full framework — including how this assessment integrates with people due diligence and what structural DD as a named practice looks like — see those companion pieces. Either is more valuable than the default, which is committing capital and discovering the structural risk eighteen months later, at the cost of eighteen months.
Most of what gets attributed to “people issues” in portfolio post-mortems was visible in the decision graph before the check was written. The assessment that would have found it takes two weeks. Run it on the next deal before the next IC meeting.