The People Who Were Not There

An enrolment exercise captures the people who were on site during the week it ran. Everybody else is handled afterwards, individually, by whoever happens to be dealing with them, and that is where the parallel systems come from.

The record in “The People Who Were Not There” becomes useful only when people understand what it proves and how to correct it. For teams exploring remote employee monitoring software, the provider website can connect time and project context with manager review, provided collection is proportionate, access is limited and consequential decisions remain subject to human explanation.

Three years later the site has a biometric clock, a drawer of paper sheets for the cleaners, a spreadsheet the agency coordinator keeps, and a set of shared codes that three people know about. None of it was decided; all of it started as a sensible improvisation during one week in March.

For an independent benchmark relevant to “The People Who Were Not There”, consult the OECD hours-worked data. Use it to test notice, accessibility, security, recordkeeping, retention and exception handling against the real operating process rather than treating a device report as self-explanatory evidence.

Who is predictably missed

People on annual leave, sick leave or parental leave during the enrolment window. Night shift, if the sessions ran during the day. Agency and contract staff, whose employer is not the one running the exercise. Cleaners and caterers on third-party contracts. Drivers and field staff who are rarely at the site. And everybody who joined the day after it finished.

On a typical site that is between a tenth and a quarter of the people who will present themselves at the reader. It is not a long tail; it is a substantial minority, and it is the minority with the least standing to complain about whatever arrangement is made for them.

Why the improvisation sticks

Because it works, in the narrow sense. The cleaner signs a sheet, the sheet gets typed in, the hours get paid. Nothing fails visibly, so nothing prompts a review.

What does not work is everything downstream. The sheet is the only record of who was in the building at ten at night. There is no audit trail on it. The person typing it in is making judgements about handwriting. And if anybody ever asks the site to produce hours worked for that contract, the answer is a drawer.

Joiners, and the week that decides it

The most common single failure is the new starter. Induction covers safety, the handbook and the tour; enrolment happens if somebody remembers. If it does not, the person spends their first fortnight borrowing a supervisor every morning, which is exactly the wrong first impression of both the system and the site.

The fix is to put enrolment on the induction checklist as a named step with an owner, and to make it impossible to complete the induction without it. That is a half-hour change to a form and it closes the largest ongoing leak.

Agency and contract staff specifically

The complication is that the organisation may not be their employer, which raises genuine questions about who may hold their biometric data and on what basis. Those questions have answers, and the answers are rarely worked out because nobody owns the problem.

The workable arrangement on most sites is to issue agency and contract staff a credential rather than enrol a biometric, with the same recording quality, the same terminal and the same audit trail. That sidesteps the data question entirely, costs the price of a card, and produces a record of equal standing. What it requires is that somebody decides it once rather than letting each coordinator invent something.

Night shift and the sessions nobody ran

If the enrolment sessions were nine to five, the night shift was enrolled by their supervisor at the end of a shift, in whatever room was free, by somebody who had watched it done once. The enrolment quality for that group is measurably worse on every site where anybody has bothered to check.

Run sessions at both ends of the clock, with the same person doing them. It costs two evenings. The alternative is a department whose refusal rate is three times the site average for reasons that look like the people and are entirely about one afternoon in March.

Finding out what you already have

One afternoon, one question: list every way a person can currently get their hours onto a payroll at this site. Ask it of payroll, of each supervisor, and of the agency coordinator separately, because they will give different answers.

Most sites find between three and six routes, of which one was designed. That list is the real scope of the system, it is almost never what the project documentation says, and it is the necessary starting point for making any of this consistent.