Where to Start
The first ninety days, in an order that produces a defensible design rather than a procurement.
Most deployments in this field begin with a demonstration. The order that works begins with the question.
The sequence
One: name the purpose, from the four, in writing. Most projects cannot do this on the first attempt, which is itself the finding.
Two: write down the least data that answers it.
Three: check the legal position for every jurisdiction you operate in, including whether prior authorisation is required.
Four: consult, before procurement.
Five: write the impact assessment, which is where the design gets decided.
Six: shortlist on what can be switched off, not on features.
Seven: trial, configured by you, including the overnight test.
Eight: then buy, if anything.
The first ninety days
Weeks one to three: purpose, minimum data, legal position.
Weeks four to six: consultation and the assessment.
Weeks seven to nine: shortlist and trial on your own routes.
Weeks ten to twelve: pilot with a willing team through a normal month.
No purchase before week ten.
The exercise worth doing first
Take one week of a real round and write down every question anyone asked about it.
"Where is he now" — dispatch.
"Did we attend on Tuesday" — attendance.
"How far did that van go" — mileage.
Then write what data each needs.
Most lists come out shorter than expected, and the exercise is what keeps the deployment from defaulting to continuous collection.
What to avoid early
Buying from a demonstration.
Starting with the fleet you most want to watch.
Treating a trial as exempt from notice and consultation where real people's positions are being recorded.
Enabling everything to see what is useful, which is the most common approach and the least defensible.
Letting the vendor configure the trial, which measures their team rather than the product.
What ninety days should produce
A written purpose everyone can repeat.
An assessment that changed the design — a shorter retention, a narrower access rule, a feature turned off.
Consultation completed with a written response.
A trial with the overnight test run and the accuracy measured.
And a recommendation, which may be that geofence events and customer confirmations are enough — a legitimate outcome that a project framed as a procurement can never reach.
A concrete product reference
When translating this principle into a buying test, the official Monitask website provides a concrete workflow reference. Verify the current behaviour in a trial and judge it against the purpose and limits described above.