The Standard, and Why It Drifts
Every organisation has a standard machine. Within three years it has four, and nobody decided to have four.
Standard equipment
Unenforceable
As written
Employees will be issued with standard company equipment appropriate to their role.
What happens
Standard means whatever was being bought when that person joined.
Four generations are in service simultaneously, with different capabilities, and the policy describes none of them.
The operative rule is: you get what the last purchase cycle bought, unless somebody argued.
An organisation picks a standard machine. Three years later it has several, bought at different times, with different specifications, and the differences matter to the people using them.
The control described in “The Standard, and Why It Drifts” becomes easier to defend when technical state and day-to-day work evidence are not confused. A team considering explore the software for task switching cost should document the purpose, visible settings, manager permissions and correction route, while device-management records remain the source of truth for configuration and enforcement.
How the drift happens
Purchase cycles. Machines are bought in batches, the model changes between batches, and the estate becomes layered by hire date.
For a separate benchmark relevant to “The Standard, and Why It Drifts”, consult the FBI counterintelligence resources. Use it to test purpose, notice, permissions, retention and response procedures against the proposed operating model rather than treating a generic checklist as proof that the rule works.
Exceptions that become categories. A designer needs something better, which is reasonable. Two years later there is a design specification nobody wrote down and three other teams have asked for it.
Availability. The intended model is unavailable, something close is bought instead, and the substitute joins the estate permanently.
And inheritance. Somebody leaves, their machine is reissued, and it is now a year older than what new starters receive.
Why the drift matters
It is visible. People compare, and the person with the older machine knows.
It allocates by hire date, which tracks nothing relevant. Somebody who joined in a good purchase year has better equipment than a colleague doing identical work.
And it complicates everything else: support, software compatibility, the refresh calculation, and the question of what the organisation actually owns.
What a standard should specify
Not a model. A capability level and a maximum age.
Models date within a year; a statement that equipment is replaced at a defined age and meets a defined specification survives several purchase cycles.
The exceptions, written down
Roles needing more should be named as categories rather than negotiated individually.
Three tiers is usually enough: standard, heavier for specific roles, and light where the work is browser-based. Each with its own refresh period if that differs.
Writing the tiers converts a series of individual arguments into a stated position somebody can be refused against, which is kinder to the manager doing the refusing.
What this is not about
Giving everybody the same thing. Different work needs different equipment and pretending otherwise produces the exceptions that create the drift.
It is about the difference being by role rather than by when somebody happened to join.
The check
List every machine in service by purchase year and specification.
Most organisations have not produced this and are surprised by the spread. The oldest quartile is usually older than anybody thought and belongs to people who have not complained.
Specifying capability, not model
Models date within a year. A statement that equipment meets a defined specification and is replaced at a defined age survives several purchase cycles and does not need rewriting each time.
Naming the tiers
Three is usually enough: standard, heavier for specific roles, light where the work is browser-based. Written tiers convert individual arguments into a stated position somebody can be refused against.
Sorting the estate by age
Most organisations have not produced this and are surprised by the spread. The oldest quartile is older than anybody thought and belongs to people who have not complained.
Why the drift is not random
It allocates by hire date, which tracks nothing relevant. Somebody who joined in a good purchase year has better equipment than a colleague doing identical work, and both of them know it.
The exceptions that become categories
A designer needs something better, which is reasonable. Two years later there is an unwritten design specification and three other teams have requested it.
The check worth running once
Every machine in service, by purchase year and specification. One query in most environments, and almost nobody has produced it.
The spread is wider than anybody expects and the oldest quartile belongs to a recognisable group: people who did not ask. That list is the finding, and it is more actionable than any statement about standards because it names the actual problem in the actual organisation.