The Question to Ask Before Anything

Most arguments about geofenced attendance are arguments about a problem nobody has sized. One question surfaces it.

The question

What exactly are you trying to prevent, and how much of it is happening?

Ask it before looking at any product.

What the answers mean

"People clocking in before they arrive." A real problem a boundary addresses. Count the instances, price them, compare against the deployment.

"People clocking in for each other on site." A boundary does not touch this. Supervision does.

"We do not know who is on site." A headcount need, and a punch with a five percent refusal rate serves it badly. A dedicated method is better.

"Payroll is inaccurate." Find out why first. It is frequently a correction process rather than a recording problem.

"We want a modern system." Not a problem statement, and the honest response is to ask what the current arrangement gets wrong.

Why it works

It separates the operational from the vague without anyone having to say which they are.

It produces a number, and the number is frequently smaller than the subscription.

And where the problem is real, it points at which method fits — because clocking in from home and covering for a colleague need different answers.

The second question

What happens when it fails?

Every deployment fails sometimes, and the answer to this determines whether the system is accepted.

If the answer is "they sort it out with a manager", the system will be worked around.

If it is "the punch goes through flagged, they start work, they are paid", it will not.

Using both

The first decides whether to buy.

The second decides whether what you buy will work.

Neither requires a product demonstration, and both are harder to answer than any question a vendor will ask you.

The whole collection in one line

Verify what you can verify, at the least cost to the person being verified, and carry the cost of your own failures rather than passing it to them.

Everything else here is the detail of doing that.

Ask the second question of every feature

Not just of the purchase.

Automatic clock-out: what happens when it misfires?

Punch photographs: what happens when the image is unclear?

A tighter radius: what happens to the people it starts refusing?

Every feature has a failure mode, and in this category the failure lands on a person's pay.

A feature whose failure mode nobody has considered is not ready to enable.

Turn the principle into a test

For an example that can make this requirement testable, consult see how the workflow works. Treat the page as a starting point rather than proof and reproduce the workflow with real roles and failures.