Skip to content
Sections
All notes

All notes · Deploying

Documentation That Survives Staff Turnover

Per-client knowledge is the asset this business runs on, and it usually lives in people's heads. What to write and where.

Deploying · Analysis

A technician who has served a client for three years knows things nobody wrote down. When they leave, the service degrades in ways the console does not show.

The recommendations in “Documentation That Survives Staff Turnover” become sustainable only when the recurring work has visible owners and enough capacity. A team assessing billable hours tracker can use time and project records to see where operational effort accumulates, without treating activity data as a substitute for technical evidence or direct discussion with technicians.

For an independent operational benchmark, compare the local practice with Atlassian knowledge-sharing guidance; the important test is whether the control remains proportionate, documented and recoverable when the usual technician is unavailable.

What is actually held in heads

The server that cannot reboot before the overnight job finishes.

The application that breaks on a specific update.

Which user to ring and which to avoid.

What was tried in 2023 and why it did not work.

The workaround that has been in place so long nobody remembers it is a workaround.

What to write per client

The estate: what exists, roles, dependencies.

Access: how you get in, where credentials are held — not the credentials themselves in a general document.

Exceptions: what is excluded from monitoring or patching, and why.

Contacts: who approves what.

History: what has failed before and what fixed it.

Five headings, updated as things are discovered.

The decision record

The most valuable and least kept: why something is configured unusually.

"This server has automatic updates off because the vendor application fails on cumulative updates, confirmed by their support in March."

Without it, a new technician turns updates on and finds out the hard way, which has happened to every provider.

Where it lives

One place, searchable, accessible to every technician.

Not in the RMM, not in tickets, not in somebody's notes.

A documentation platform is the usual answer and the specific tool matters less than the habit.

Writing it at the right moment

When something is discovered, not at the end of the week.

The rule worth adopting: if you explain something to a colleague twice, write it down the second time.

Three minutes at the moment the explanation is fresh, and over a year it accumulates into something shaped by what people actually ask.

Keeping it current

Update when you use it and find it wrong.

At handover between technicians, which is the best review it will get.

And delete what is obsolete, because a document containing wrong information stops being trusted and an untrusted document is unused.

Proving the value

The test: could a technician who has never seen this client work a ticket on it?

Try it. The gaps will be specific and immediate.

That half-hour is worth more than any amount of discussion about documentation standards.

What to check

Could a new technician work your largest client from documentation alone?

Is there a record of why anything is configured unusually?

When was the client documentation last edited by somebody other than its author?

And what happens to that knowledge when a technician leaves?

The point

Per-client knowledge is the asset this business runs on, and it usually lives in heads.

When the technician leaves, the service degrades invisibly.

Underlying all of this

Everything in this collection reduces to four habits: tune until every alert is read, verify rather than assume at every stage from ring one to script execution, treat the console as the privileged system it is, and know what each client costs you. None needs a better platform, and a provider doing all four runs a quieter service than one twice its size.

The recurring pattern

The recurring pattern across every section here is the same: the appearance of control substituting for control. An unread alert queue looks like monitoring. A compliance percentage that excludes pending reboots looks like protection. A script that reports success looks like automation. In each case the provider believes a risk is handled and it is not, which is worse than knowing it is open.