What leaves the operation when your most experienced person does

The most experienced person on your team carries more of the operation in their head than anyone wants to admit. Not the documented version. The real one. Which supplier delivers late on Fridays. How the biomass numbers should be read before they are entered. What a specific exception in the shift log actually means versus what it looks like on paper. This person is invaluable and almost entirely invisible in the org chart.

The operational knowledge problem

When that person leaves, the operation does not stop. It becomes noisier. Decisions take longer. The same question gets asked in three different places. A compliance report that used to take two hours takes four, because nobody is sure what the shortcut was, or why it existed. The team is capable. The knowledge just left with one person.

In Norwegian operations, particularly aquaculture, maritime services, and logistics across Trøndelag and the coast, this pattern is common and underestimated. The work is complex. The teams are lean. Turnover, however low, has outsized impact when institutional knowledge is concentrated in a handful of people. An experienced site manager carries years of local pattern recognition. When they go, that does not transfer automatically.

The real cost is harder to measure than a recruitment fee. It lives in slower decisions, more errors during transitions, compliance gaps that are discovered rather than anticipated, and a growing distance between how the operation appears to perform and how it actually runs day to day. For operations with Mattilsynet oversight or strict biomass reporting requirements, this gap is not just inefficient. It is a liability.

What structured knowledge capture actually looks like

The instinct is to write a manual. Manuals are the wrong format for the kind of knowledge that matters here. They capture the policy version, not the operational one. Nobody writes down the things they do automatically. They write down the steps they think a new person would need, which misses most of what actually matters.

AI-assisted knowledge systems work differently. They pull from actual workflow data: shift handovers, incident logs, compliance records, maintenance notes. They build a queryable layer of operational context that any team member can access when needed. When a new site manager needs to understand why the morning check runs in a specific sequence, the system draws on hundreds of previous handover records to explain the pattern. Not a policy document written once and never updated. The actual reasoning, surfaced from real work.

For aquaculture operations on Frøya and Hitra, the most practical starting points are shift handover and compliance documentation. These are the highest-frequency knowledge transfer moments and the most likely to carry undocumented context. Building structured capture around these two touchpoints creates a retrievable layer that survives staff transitions.

In logistics and multi-site coordination, the equivalent is route logic, exception handling, and supplier relationship context. What an experienced dispatcher knows about which routes run long, which supplier needs a phone call instead of an email, and what the early signs of a delivery problem look like. This knowledge evaporates quietly and costs real time on the other side.

Starting point

The right entry point is not a knowledge management project. Those take months to scope, cost significant budget, and usually produce a wiki nobody updates. The better approach is a single handover.

Take the most knowledge-dense recurring transition in your operation: the shift handover, the weekly status report, the end-of-season compliance review. Build structured capture around that one moment. Not a form, but a system that extracts structured information from what your team already writes and makes it findable and retrievable.

When you can pull up the last fifteen times a particular exception occurred and see what the team did each time, you have built something that survives turnover. You have also built the foundation for everything else: faster onboarding, more reliable compliance, and decisions that do not depend on one person being in the room.

Start there. Prove it works on one handover. Expand from there.

If you want to know what this looks like for your operation, start here.