Your process map is a geological record of every failure the organisation has survived. Each layer is a workaround for a failure mode that may no longer exist, and the people who navigate the layers are the ones who would lose the most if you removed them. That’s not an efficiency problem. It’s structural gravity.
Climate tech organisations accumulate this faster than most. The coordination load is higher — hardware and software, government and commercial, project and product — and the failure modes leave bigger scars. A missed milestone on one municipal pilot becomes a three-week review loop on every future pilot. A partial shipment to one reinsurer customer becomes a five-signature approval on every future shipment. Each workaround is individually rational. Nobody ever removes one after the underlying condition changes.
The trap isn’t that the process is complicated. The trap is that nobody in the organisation can distinguish the layers that are load-bearing from the layers that are residue from problems nobody remembers.
Adaptation compounds in one direction only
Organisations adapt. That’s not the problem. The problem is that adaptation compounds in one direction and almost never in the other. A customer escalation adds a review step. A production issue adds a sign-off. A key person leaves and the knowledge loss gets patched with a documentation requirement. Each addition is defensible in isolation. The subtraction that should follow — “this review step is no longer needed because we redesigned the handoff” — almost never happens, because the person who added the step has moved on and nobody remembers what the step was originally preventing.
Every time a layer gets added without an old one being removed, the organisation’s coordination cost rises non-linearly. A ten-step process that grows to fifteen steps isn’t 50% slower. It’s three to four times slower, because every added step multiplies interaction effects with every other step. The company isn’t slow because the work is complex. The team is carrying fifty kilos of accumulated layers, ten of which are load-bearing, and nobody can tell which ten.
This is why accidental complexity is the sibling pattern to the adaptation trap. Accidental complexity is what the founder’s cognitive model generates on day one. The adaptation trap is what accumulates on top, layer by layer, year by year, long after the founder stopped being the only person adding layers.
The people who navigate the workaround are the ones who defend it
The reason the trap is hard to remove isn’t that nobody can see it. It’s that the people who can see it are the ones who’ve built expertise in navigating it, and that expertise is real. The senior operations lead who knows which review steps can be bypassed with which approvers is genuinely creating value inside the current system. Simplifying the process devalues their skill. They will resist hardest, and their resistance is not irrational — it’s self-preserving.
This is the specific reason that “process redesign” as an intervention usually fails. The redesign replaces the adapted process with a new clean process. Within eighteen months, the new process has accumulated its own layers, because the same people are still running the same organisation under the same structural conditions. The trap reproduces one abstraction later. What changed was the label, not the underlying dynamic.
The diagnostic signal is the ratio of actual to official
The diagnostic question isn’t “is this process too complex?” Every founder will tell you yes. The diagnostic question is: what’s the ratio of the actual process to the official process? Map a real piece of work from initiation to completion, including every workaround, every side conversation, every “just slack so-and-so” shortcut. Then put the official process next to it and count.
A 1.5x ratio is normal. Every organisation has some drift, and the drift is healthy because it adapts around edge cases that formal process can’t anticipate. A 3x ratio means the official process is fiction and the real work is happening through a shadow network of relationships. A 5x+ ratio means the company’s operating capacity is being held together by a small number of people whose tribal knowledge has become a single point of failure. When those people leave, the company doesn’t slow down — it stops, and nobody can rebuild the shadow process from the documentation, because the documentation was always the fiction.
I watched this pattern from inside a climate tech company building adaptation products. The sales-to-delivery process was documented as four steps. The actual process had eleven, and six of them were personal relationships with specific implementation partners that nobody had formalised. When two of those partners moved on, onboarding time tripled — not because the new partners were worse, but because the shadow process had evaporated and nobody had a documented replacement. The company had been running on tribal coordination for two years and calling it “process.”
What changes isn’t the process, it’s the archaeology practice
The intervention isn’t another process redesign. Those produce a clean new process that the organisation immediately begins adapting around. The intervention is building the discipline of process archaeology: a quarterly review that asks, for each process step, “does this exist because of a current operational requirement, or because of a historical failure that may no longer be relevant?” Most steps can’t survive that question honestly asked. The ones that can are the load-bearing ones.
The goal isn’t simplicity for its own sake. The goal is a process that reflects current operational reality instead of the accumulated history of every failure the organisation has survived. For portfolio companies underperforming against thesis, the adaptation trap is often the invisible variable — the company isn’t slow because the team is bad, it’s slow because the accumulated process weight makes speed structurally impossible. For post-investment support, mapping the actual-vs-official ratio is the first assessment to run, because it’s the cheapest diagnostic that reveals the largest structural problem.
If your team’s velocity has dropped and nobody can name what changed, the change was probably gradual — a new approval here, a new review there, a new “just to be safe” that never got removed. Map the actual-vs-official ratio before the next planning cycle.