Every record has two readers: the colleague tomorrow, and the reviewer someday. Good documentation serves both with the same sentences.
Every record made in a public-serving organization has two readers. The first is known and immediate: the colleague on the next shift, the worker who inherits the case, your own future self trying to remember why. The second is hypothetical and distant: the auditor, the ombudsman, the coroner, the court — the reader you hope never comes, who arrives years later knowing only what the file says.
Most documentation guidance treats the second reader as a threat, and the writing it produces shows it: defensive, voluminous, engineered to demonstrate that process was followed rather than that thinking occurred. This is a mistake about what scrutiny actually asks. The reviewer who matters is rarely asking whether the organization was perfect. They are asking whether the people involved thought carefully with the information they had. A record can only answer that question if it contains the thinking.
Which is why writing for the second reader, done properly, is not defensive practice at all. It is the same discipline as good practice: what was decided, on what information, considering what alternatives, for what reason. A file that would satisfy a coroner is, almost always, a file that would genuinely help the colleague on the next shift. The two readers want the same thing. Only bad documentation strategy separates them.
The common failure modes are worth naming. The first is the outcome-only record: the decision is noted, the reasoning is not. Under scrutiny, absent reasoning reads as absent thought — and the worker is left defending a years-old judgement from memory, which protects neither them nor the organization.
Under scrutiny, the reasoning that was never written does not exist.
The second failure is volume as safety — the belief that more fields, more forms and more mandatory sections make an organization more defensible. They usually do the opposite. The signal drowns; the hours that the records which matter deserved are spent on records that protect nothing; and staff, unable to say who any given form is for, stop believing that any of it matters.
The third is the undocumented exception. Practice sometimes departs from policy for good reason — the policy did not foresee this case, and the worker judged well. Recorded with its reasoning, that departure is defensible judgement. Unrecorded, the same departure discovered later looks like carelessness or concealment. The rule is simple: the better the reason for the exception, the more important it is to write the reason down.
What does this ask of leaders? Three things, mostly. Set the standard that consequential decisions carry their reasons — options considered, information relied on, why this course — and supervise to it. Retire the recording that serves no reader, deliberately and on someone's authority, so the time exists to do the recording that does. And talk about documentation as what it actually is: the organization's memory of its own judgement, and the only version of events that survives staff turnover, fading memory and time.
A record's job, in the end, is to let someone who was not there conclude that the person who was there thought carefully. Files that do that job protect everyone in the chain — the person served, the worker, the supervisor, the organization. Files that merely prove process was followed protect the process.