What happens when someone on your offshore team leaves: designing for continuity

Every seat turns over eventually. A person gets promoted, moves, or changes direction, and one day the work they were doing needs somebody else to do it. That is an ordinary event in any company, onshore or offshore.

What varies enormously between companies is the next morning. In one company the departure is a transition. The work continues, the standard holds, and whoever picks the role up starts from a role that was written down before anyone filled it. In another company the same departure is an excavation. People spend weeks working out what the person actually did, which recurring tasks nobody else knew about, and where the current version of anything is kept. Same event, two very different bills. The difference had almost nothing to do with the person who left.

Why one departure costs more than another

A departure costs in proportion to how much of the work only ever existed in the way one person did it. That is the whole variable, and the mechanism behind it is more sympathetic than it sounds.

Nobody hoards. What happens is that a capable person is handed a role with real ambiguity in it, and they resolve the ambiguity themselves, because the work has to ship. They invent the sequence. They decide what good enough means on a Thursday when there is nobody to ask. They build the workaround for the system that does not quite do the thing. Every one of those decisions is the company getting served, and it is a large part of why the quarter held. Every one of them is also a piece of the operating method that now exists nowhere except in their practice.

So the exposure accumulates quietly, built by good people doing the right thing under ambiguity. It stays invisible while they are there. It presents itself all at once on the day they are not, which is why it arrives looking like a personnel event when it is really a design bill turning up late. The pattern is familiar. It is the same one that decides whether an offshore arrangement works at all, showing up at the exit instead of at the start.

A written seat is the first thing that survives

The first thing a company still holds the morning after is the role itself, if the role was ever written down. A defined seat records what the work owns, what it excludes, what done looks like against a real example, the rhythm it runs on, and where the current truth of the work is kept. In Operational Alpha that page is the Seat Card, and how to write one before you hire is its own article.

The page is written to make the hire work. Its second job only shows up later. Whoever takes the role next reads the answers instead of hunting for them, and starts from the role. Without the page, that same person starts from an investigation, and the investigation gets conducted by interviewing everyone who sat near the work, which means the cost of one departure lands on the senior people around the seat rather than on the seat.

Notice which way the incentive runs, because this is where most continuity advice goes wrong. Documentation written for its own sake decays. Nobody reads it, so nobody maintains it, and within two quarters it describes a job that no longer exists. A seat definition stays current because the seat is run from it. Continuity turns out to be a by-product of running the work off a written definition, which is exactly why it is still accurate on the one day it is needed.

The team is the unit of continuity

The second thing that survives is whatever the rest of the team already saw. A seat that runs alone has nobody to hand to, and that is what a single point of failure actually is: one person, one channel, and nobody else who ever saw the work while it was being done.

An operator who works on the same rhythm as the rest of your team is seen by other people in the ordinary course of the week. Colleagues watch the reasoning take shape rather than only the finished output, so the method stops being private long before anyone thinks to write it down. That shared sight pays twice. It is what lets a team catch a performance problem early while the person is still in the seat, and it is what leaves somebody who can carry the work when the person is not.

Above the team sits the layer that owns role health, and that is where continuity accumulates on purpose. Somebody outside the seat knows what the role has become since it was written, which parts of it drift, what the last three months actually looked like. That is a second copy of the role's context, held deliberately by design rather than by accident. It is the part of the interface around a seat that a departure tests hardest.

What you still hold the morning after

Three things decide the morning, and all three are set in advance.

• The role, on a page. What the seat owns, what it excludes, what done looks like against a gold example, and where the truth of the work is kept. The next person starts from the definition instead of an investigation.
• The work, inside systems you already own. If the work was done in the company's own systems, the output and the history of it stay put when a person goes. The access half of the same question, including revocation you can execute yourself, is a separate list worth requiring in writing.
• The context, held by more than one person. The team saw the work in progress. The stewardship layer knows what the role has become. Neither of those is a document, and both of them survive a departure.

A departure does not create the exposure. It reveals where the design was thin, on a day nobody chose.

Where Kayana fits

Everything above describes work done before anyone needs it. Kayana creates Mini-GCCs for PE-backed middle-market companies. The client owns the Seat and defines what done means, and we design and run everything around it, which is how both halves of a transition get settled long before one arrives. Every seat we place starts from a written definition, and our operators work inside the client's own systems. The strongest continuity design is still the one that gets used least often. Voluntary turnover across the teams we build runs 3–5% in year one, which means fewer of these mornings to absorb in the first place.

Quick answers

The design system behind all of this is the book. Operational Alpha →