Her best department head left. The successor spent nine months rebuilding.
...what had never been written down.
A communication agency I know well had launched, a few years back, a new content production line for its key accounts. The head recruited to build it had set up everything: the rate card, the way services were packaged, the trade-offs on internal resources absorbed without being billed to the client, the calibration of margins differentiated by deliverable type, the organization of a team of writers and art directors working remotely.
It worked. The activity was profitable. The department ran.
After a few years, the head left for a competitor. Replacement was fast; a senior profile, experienced, recruited on his editorial track record and his managerial credentials. The org chart was rebuilt within weeks.
The successor then spent nine months understanding how the activity actually worked.
What those nine months cost
Not the rate card: it was in the CRM and in commercial proposals. Not the official processes: they were documented in the quality manual.
What had to be reconstructed over nine months was what had never been written. Why a given margin was higher on a given deliverable format, and why that gap was structurally necessary. Which creative resources were absorbed internally rather than billed to the client, under what logic, with which external suppliers and on which negotiated terms. How each writer and each art director, scattered across several cities, actually organized their work. Why some operational trade-offs that looked inconsistent on paper were in fact what produced the profitability.
Through those nine months, the successor made decisions on a partial understanding. Some choices reproduced by chance what his predecessor had done; others undid invisible balances that had to be restored later. The department’s profitability fluctuated during this period without anyone — successor or leadership — being able to attribute the variations to specific causes.
None of these losses were accounted for. None appeared in the replacement budget.
What this story reveals about service SMEs
The critical knowledge of a service SME does not reside in its procedures. It resides in the heads of those who built its activities. Contracts, job descriptions, org charts, quality processes capture a useful but largely insufficient fraction of how the firm actually works.
Nonaka (1994) drew the distinction that frames this observation thirty years ago. Explicit knowledge — codifiable, transferable through documents — is only one part of an organization’s cognitive capital. Tacit knowledge — embedded in habits, intuitions, daily trade-offs — is the other part, and it is precisely the one that produces a service SME’s competitive edge. Its conversion into explicit knowledge never happens spontaneously; it requires intentional mechanisms that few organizations install.
Recent research on SMEs shows the gap between awareness and action. A European study (JEIM, 2023) estimates the average time to rebuild critical knowledge lost through an unprepared departure at 18-24 months. A Polish survey of 380 SMEs (2024) puts at 83.7% the share of firms underinvesting in systematic knowledge capture, despite a stated awareness of the risk. The gap between recognizing the problem and installing mechanisms is large, and it is large because the cost of non-capture stays invisible until the departure.
Why knowledge stays locked in
Knowledge retention is not, most of the time, the result of individual ill will. It comes from several reinforcing mechanisms.
First, tacit knowledge does not know itself. The person who holds it does not perceive what is obvious to her as transmissible knowledge. “Of course we absorb that cost internally, it’s logical”; except that the logic in question rests on a chain of historical reasoning and trade-offs that no document captures.
Second, remote organization amplifies opacity. When the team is distributed, the head no longer directly observes how each contributor works. The rituals that once produced transparency — coffee breaks, hallways, open-plan offices — disappeared without being replaced by equivalent mechanisms. The successor I described in the introduction inherited not only uncaptured knowledge but a team whose daily functioning was legible to no one outside each of its members.
Finally, and this is the most delicate part, in teams where trust between contributors is low, each person preserves what she knows. The phrase “I work the way I want and I share nothing because the others are incompetent” is not a caricature; it is an observable stance, and it is individually rational in a context where sharing exposes you to criticism without counterpart. Knowledge then becomes a private asset that each contributor protects, with or without intent to harm the organization.
What classical documentation does not capture
Most service SMEs confuse capture with documentation. They install a wiki, a SharePoint, a DMS; they ask teams to “document”; they end up with procedure sheets that describe the what without capturing the why.
What the successor lacks is almost never the what. He can deduce it from contracts, invoices, briefs, past deliverables. What he lacks is the reasons for the trade-offs: why this margin, why this supplier choice, why this writer on this client, why this departure from the standard rate. Reasons live in historical decisions that accumulated without being traced.
Classical documentation answers a question no one asks at a departure. The successor does not ask how the process works; he asks why it works the way it does.
What could have been installed beforehand
Three complementary mechanisms separate organizations that survive a key departure from those that spend nine months rebuilding.
The first is a decision log kept by the head of an activity, in a light, regular format. Not meeting minutes; a trace of non-trivial trade-offs along with their rationale. When you choose one supplier over another, when you absorb a cost internally, when you adjust a margin, when you depart from the standard rate, you record the decision and you record the reason. After a year, this log is worth more than the quality manual.
The second is a regular ritual — monthly, quarterly — where the team of an activity collectively articulates how it actually works. Not how it is supposed to work; how it does. This practice produces two effects: it converts tacit into explicit as work evolves, and it reveals to the team itself what it did not know it knew.
The third is the deliberate development of a deputy (”adjoint”) for every critical role. Not an administrative backup, not an occasional delegate; a true deputy who participates in structural decisions, knows the suppliers, understands the margins, and can hold the role through an extended absence.
In my management courses, I advise every manager to identify and develop a deputy, regardless of departure risk. The reason is not only continuity; it lies in what the exercise produces in the manager herself. Developing your deputy forces you to articulate what you do, why you do it, and how you make the trade-offs you used to make on instinct. It is one of the rare practices that forces a manager to understand her own organization, and it is probably the best training in delegation. Knowledge capture is its natural by-product; organizational resilience to a departure is its direct consequence.
The cost of developing a deputy is easy to quantify; the cost of nine months of reconstruction almost never is.
The capture test
One question is enough to place an organization. If the head of a given activity left tomorrow with no notice, how many months would it take a competent successor to reach 80% of current performance? If the honest answer is between six and twelve months, the activity depends on uncaptured tacit knowledge. If it goes beyond twelve months, the activity is structurally fragile. Most service SME leaders I ask answer “between nine and fifteen months” without hesitation; and yet draw no operational consequence from their own answer.
The question to ask this week
Identify the three most profitable activities in your firm. For each one, ask yourself what share of the critical knowledge is captured today in a form a competent successor could absorb in less than three months. If the answer falls below 50% for even one activity, you now know the most profitable project you have not yet launched.
Lionel Jaquet — CVO @tebicom · Responsive Management · DBA Candidate @GEM