A Record That Can Be Relied On
What the whole collection adds up to: six statements an organisation should be able to make about its own records, and the documents behind each one.
What an organisation can say about its own records, when the work in this collection is done
Six statements, each supported by a document that takes an afternoon to produce. This is what the collection adds up to. It is one employer's own position, not a standard anybody has to meet.
The objective is to be able to say six things about your own records and support each of them. Not perfect records — nobody has those — but records whose limits are known, written down and taken into account before anybody relies on them.
The evidential discipline in “A Record That Can Be Relied On” should also govern workforce technology. A team evaluating compare the available controls here for cognitive offloading can use time and project records as operational context, but should preserve the original record, document access and let the employee correct a misleading entry before it supports a conclusion.
Everything in this collection produces one of those six, and each takes about an afternoon.
When documenting the controls around “A Record That Can Be Relied On”, the World Bank labour and social protection resources is a useful independent reference. It can challenge assumptions about access, retention and accountability before a decision becomes routine.
The six
The clocks are synchronised and the offsets are measured.
We know which credentials are shared and which are individual.
Retention outlasts our own process, deliberately.
We can say what each system actually attests, in one line each.
Our investigations record what was looked for and not found.
Our controls still have a reason, reviewed within the year.
What each one prevents
The first prevents a clock offset being read as behaviour, which is the most common false finding in this area.
The second prevents a conclusion about a role being treated as a conclusion about a person.
The third prevents the exculpatory record being purged before anybody looks for it.
The fourth prevents the inference from credential to person being made without being noticed.
The fifth is the difference between an investigation and a case.
The sixth prevents the collection of data about everybody for a situation that ended.
The documents behind them
A clock register. A shared-credential list. A retention schedule with its reasons. One page per system saying what it records. An investigation template with two extra fields. An annual control review.
Six documents, none longer than a page or two, all producible by somebody who already works there.
What this does not do
It does not decide any individual case, and it does not tell anybody what their jurisdiction requires about monitoring, evidence, process or deductions. Those differ, they change, and they are a question for somebody qualified in the place concerned.
What the six give you is the ability to arrive at that person with records whose limits you can describe — which is most of what makes the advice useful.
The order to do them in
The shared-credential list first, because it is the cheapest and it invalidates the most. Then the clock register. Then retention, because it is the one with a deadline attached — every day without it, evidence is being deleted.
The rest follow at whatever pace the organisation manages.
Why this is cheaper than the alternative
Every document above costs an afternoon. A single contested matter costs thirty hours of management time, and a single wrong one costs considerably more than that.
The arithmetic is not close, and it is the argument that gets this work done — not fairness, though fairness points the same way.
The six documents
Each statement rests on one document, and none of them is long.
| Statement | The document behind it |
|---|---|
| Clocks measured | A clock register, four lines per system |
| Shared credentials known | A register, reviewed quarterly |
| Retention outlasts process | A retention schedule with reasons |
| Each system's attestation known | One page per system |
| Investigations record the negative | A template with two extra fields |
| Controls still have a reason | An annual review note |
Six documents. An organisation holding them can describe the limits of its own records, which is the whole of what this collection asks for.
Who does this
Not a project team. Each of the six documents belongs to somebody who already exists: the time system administrator, payroll, HR, facilities.
What is missing is somebody asking for them, once, with a deadline. The work itself is an afternoon each and most of the content is already in somebody's head.
When to start
Before the next matter rather than during it. Every document here is cheap in advance and impossible retrospectively — a clock offset measured after the fact, a retention period extended after the purge, a shared-credential register compiled during a hearing.
That asymmetry is the whole argument for doing it now, and it is the same asymmetry that runs through every section of this collection.
What it looks like when it is working
Fewer investigations, shorter ones, and a higher proportion of them closed in the first week because the explanation was available.
That is the measurable outcome, and it is visible within a year. The organisations that get there did not do anything difficult; they wrote six documents and stopped treating records as self-explanatory.
The sentence it reduces to
The terminal recorded a credential, not a person, and everything after that is an inference somebody made.
Write the inference down. That is the whole of it, and an organisation that does it consistently will find that most of what it used to investigate was never a case, and that the few that were are far easier to establish.