Choosing a System
The questions that separate a product you can configure down from one designed to collect everything.
Most trackers do the same thing technically. They differ in what they let you switch off.
The configuration questions
Can continuous position logging be disabled while keeping geofence events?
Can collection be suspended outside working hours, automatically, and how is the boundary determined?
Is there a driver-operated pause, and does it stop transmission rather than just hiding the data?
Can retention be set per data category, or is there one global number?
Can distance be reported without storing the route?
A product that answers no to most of these is designed to collect everything, and no access policy compensates for that.
The worker-facing questions
Rarely asked in procurement and they decide how the deployment is received.
Can a person see their own data?
Can they see when their position was read, and by whom?
Is there a visible indicator when collection is active?
Can they export their own record?
These exist in better products and are absent from most. Asking moves the conversation immediately.
What to check in the account
The defaults, before changing anything. What is switched on at account creation tells you what the product is for.
Whether the map view is permanent or tied to an operational action.
Whether read access is logged.
What happens after a vendor upgrade, because features arrive enabled.
Red flags
Marketing about catching people rather than about running an operation.
"Stealth mode" or any feature describing concealment.
Driver league tables as a headline capability.
No pause control, or one the administrator can override invisibly.
Retention defaulting to six months or more.
Vagueness about where data is processed.
The trial
Configure it yourself, because after go-live you are the one configuring it.
Drive a known route and compare against ground truth — this tells you the real accuracy.
Test a geofence at a real customer site, including approach and departure.
Test the pause: switch it on and confirm nothing arrives.
Leave a vehicle parked overnight and check what the system recorded.
That last test is the one that matters most and almost nobody runs it.
The contract
Processing location and subprocessors, named.
Retention matching yours, with deletion confirmed on termination.
Notification before feature changes that alter what is collected.
Export in a readable format, specified now rather than requested at the end.
Check the defaults on day one
Before configuring anything.
Create a trial account and list every capture feature and its state.
Document it, because that starting state is what tells you what the product is designed to do.
Then configure down, and record what you turned off.
Compare after every upgrade, which is the only way to know the exclusions still hold.
A concrete product reference
When translating this principle into a buying test, employee monitoring software pricing provides a concrete workflow reference. Verify the current behaviour in a trial and judge it against the purpose and limits described above.