Controls That Create New Problems
The second-order effects of time controls: queues, unreviewed data, personal devices, unpaid supervision, and the control that makes the behaviour worse.
Five controls and the problem each one introduced, at one employer
Four of the five controls worked, the fifth stopped nothing, and every one of them created a second problem the business case did not mention. This is one employer's own account, not a statement of what any control is permitted to do where you are.
Every control works, and most of them create a second problem the business case did not mention. Biometrics end credential sharing and double the queue. Photo capture catches proxy punching and generates twelve thousand images a month that nobody looks at. A tighter lateness rule produces more badge-passing, not less.
The control described in “Controls That Create New Problems” should be matched by transparent operating rules. Organisations considering review the evidence-handling options for time tracking with screenshots can make that use more credible by publishing the purpose, selecting only necessary settings, limiting manager access and fixing a review date before the first record is collected.
None of that is an argument against controls. It is an argument for writing down what each one will produce as well as what it will stop, before buying it.
The ISACA governance resources offers another lens on the issue raised in “Controls That Create New Problems”. Compare its principles with the actual record, ownership model and review route rather than importing a generic checklist unchanged.
The four second-order effects
Queueing. Any control that adds a step at the terminal adds it for everybody, every shift, and the arithmetic is unforgiving: four seconds per person across sixty people is four minutes of queue.
Data nobody uses. Photos, logs and footage that are captured and never reviewed are an obligation without a benefit — they have to be stored, secured, retained and eventually disclosed, and they catch nothing because nobody looks.
Shifting the cost onto people. Geofenced apps that require a personal phone, biometric terminals that require a body that reads, supervisor sign-off that requires an unpaid early arrival.
Making the behaviour rational. Controls that increase the penalty without removing the cause increase the incentive to avoid the system.
Reviewing what gets captured
A control that produces a record implies somebody reviewing it. If the review has no owner, no schedule and no criterion, the capture should be reconsidered rather than the review added to somebody's list informally.
- Who reviews, by role?
- On what schedule, and against what criterion?
- What happens when something is found?
- How long is the material kept, and who may see it?
- What is the plan when nobody has reviewed anything for three months?
The last question is the honest one, because that is the actual future of most capture systems.
The control that makes it worse
Tightening a lateness rule in response to badge-passing is the intervention most likely to backfire, because the rule is usually the cause.
That is easy to test and almost never tested: change the rule for a month on one shift and measure the clustering. A result either way is worth more than an opinion, and it costs nothing.
Proportionality as an operational question
Separate from whatever the law requires where you operate — which is a question for somebody qualified in the place concerned — there is a plain operational version: is what this control collects proportionate to the problem it addresses?
An employer that quantified the problem first can answer that. One that reached for a control on the strength of an impression cannot, and the inability to answer is itself a risk.
Telling people what changed and why
A control introduced without explanation is read as an accusation, and read that way it damages exactly the cooperation that makes the lighter alternatives work.
The explanation does not have to be elaborate: what the problem was, what was considered, what was chosen and why. A workforce that was told the terminal queue was the problem and sees a second terminal appear draws a different conclusion from one that finds a fingerprint reader in the morning.
Piloting on one shift
Nearly every control described here can be tried on one shift or one site for a month before being bought for everybody.
That is cheap, it produces a real number rather than a vendor's, and it surfaces the second-order effects while they are still reversible. The reason it rarely happens is that a pilot delays a decision somebody has already made — which is precisely when a pilot is most useful.
The support burden
Every control adds a stream of small failures that somebody has to handle: a reader that will not read, an app that will not open, a photo that did not capture.
That load is real, it falls on supervisors, and it is absent from every business case. Asking the vendor what support volume to expect, and believing the answer only after a pilot, is the way to find out before committing.
Reviewing a control you already have
Most organisations inherit controls rather than choose them, and nobody revisits them. A review is short: what does it stop, what does it cost, what does it collect, who reviews that, and does the original problem still exist?
Five questions a year, per control. It regularly finds something that was introduced for a situation that ended, is still collecting data about everybody, and is reviewed by nobody — which is a worse position than having no control at all.