Enrolling More Than One Thing

The single cheapest improvement available to a biometric time clock is to enrol more than one thing per person: several fingers, two modalities, or a biometric plus a card. It costs well under a minute each and it removes a large share of everything described in the rest of this collection.

The enrolment decision in “Enrolling More Than One Thing” is the moment to explain the whole attendance workflow, not only the reader. An organisation considering how to monitor employees without being intrusive for how to monitor employees without being intrusive should tell people which records managers can see, how time can be corrected and what fallback applies before any optional identity or activity setting is enabled.

It is skipped almost everywhere, for an understandable reason. The enrolment session is a queue of people taken off their work, the software reports success after the first capture, and nobody present has any reason to think the extra capture will matter. The people who find out that it mattered are working a different shift, three months later.

For an independent benchmark relevant to “Enrolling More Than One Thing”, consult the CIPD working-time guidance. 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.

Multi-finger enrolment

Enrol at least two fingers on each hand where the device allows it, and record which. A device permitted to try any enrolled finger will find a readable one on most of the mornings when the usual finger has a plaster on it, has been in solvent, or has simply gone cold.

The improvement is largest exactly where the need is greatest. Among manual trades, single-finger enrolment commonly leaves one person in eight needing a fallback regularly; three-finger enrolment typically cuts that by two-thirds. The arithmetic is unusually clean: half an hour of extra enrolment time across a hundred people against several hundred fallback events a year.

A second modality, where it exists

Some systems support face and finger on the same terminal, or finger and hand. Where that is available and the licensing does not punish it, enrolling both gives a genuine alternative rather than a degraded one, because the second route is still a biometric and still records the person themselves at the moment they arrive.

Where it is not available, the second route is a card or a code, which is fine. The point is not purity of method; it is that something works at the terminal, immediately, without another human being.

A card for everybody, including the people who do not need one

This is resisted on cost and it is usually the right call anyway. A credential costs a few units of currency; a fallback event costs several minutes of two people's time.

It also solves the problems that have nothing to do with biometrics: the reader that is down, the network that is out, the new starter enrolled this afternoon, the visitor from another site. A site where everybody has a card and most people never use it is a site where an outage is an inconvenience rather than a morning spent writing names on a clipboard.

Deciding the second method per person, at enrolment

The line that should exist on every enrolment record and almost never does: for this person, if the reader refuses, they will use X, authorised by Y, recorded as Z.

Filling it in takes five seconds for the ninety per cent who will never need it and is the whole arrangement for the ten per cent who will. It also forces the question to be answered at the one moment when the right people are in the room — the person, somebody who knows the system, and somebody with authority — rather than at six in the morning when none of them are.

Testing the second route on the day

Having enrolled a backup, use it once, there and then. Thirty people presenting their card or entering their code at the live terminal, an hour after enrolment, finds the configuration problems that otherwise surface one at a time over six weeks.

The common findings are mundane and worth finding early: codes issued but not activated, cards enrolled on the access system and not the time system, a terminal that accepts a code but records it as an override rather than as a normal entry. Each takes minutes to fix on the day and days to chase afterwards.

What it costs, stated honestly

Thirty to sixty seconds more per person at enrolment, a card for everybody, and one additional line in the enrolment record. On a site of four hundred people that is about five extra hours in total and the price of four hundred cards.

Against that: several hundred fewer fallback events a year, each consuming a few minutes of two people, plus the disputes, the overrides and the quiet resentment that do not appear in any figure. It is the clearest cost case in the whole subject and it has to be made before the enrolment week, because afterwards nobody will reconvene four hundred people to do it properly.