Sizing Before Shortlisting

Vendors quote per seat. The bill is decided by things that are not seats, and knowing them early removes half the shortlist for free.

The four numbers

People recording time, now and in three years. Seasonal peaks matter: some pricing counts named users, some counts active users in a month, and the difference across a peak can be double.

Entries per person per week. A desk worker recording four projects a day generates twenty entries a week; a field worker clocking in and out generates ten. Some products slow noticeably at high entry volumes, and some price on them.

Distinct rule sets: jurisdictions, union agreements, contract types, pay grades with different overtime treatment. This drives the configuration effort more than any other figure.

History to migrate, in years and in rows, including the audit trail.

What each one eliminates

Headcount decides whether you are in the self-service tier or the enterprise one, and those are different products with the same name.

Entry volume eliminates products whose interface assumes a handful of entries and whose reporting times out above a threshold. Ask for a reference at your volume.

Rule sets eliminate products with a fixed rule engine. A product that handles overtime as a single setting cannot serve three jurisdictions, and no amount of configuration fixes that.

History volume decides whether migration is an import or a project, and whether you can keep the old system readable instead.

The numbers vendors ask for and the ones they should

They ask headcount, because that is the price.

They should ask rule sets and entry volume, because that is the implementation.

A vendor who asks only headcount is quoting for a licence, not for a working system, and the implementation cost will arrive later as a change request.

Getting the numbers

Headcount from payroll, split by worker type.

Entry volume from the current system, or by sampling: ask ten people to count their entries for a week.

Rule sets by listing them, which is an afternoon and which frequently finds rules nobody had written down.

History by querying the current system, in rows rather than in years.

Where sizing goes wrong

Counting only employees and forgetting contractors, agency staff and seasonal workers who also record time.

Assuming growth is linear when a planned acquisition doubles it.

Ignoring the peak, which for seasonal businesses is the number that matters.

Treating history as optional. Statutory retention means the old records must remain accessible for years, and deciding how late is expensive.

Where the tier boundaries fall

A question that changes the shortlist and is rarely asked.

Ask each vendor where their pricing tiers change, in seats and in volume.

Compare against your three-year projection.

A product that is cheapest today and crosses a boundary in year two is not cheapest.

Ask what happens at the boundary: a price step, a different contract, or a migration to a different edition of the product. The third exists and is the expensive one.

Model the real operating unit

Seat totals alone miss clients, temporary workers and permission groups. Use one industry example to identify those dimensions before asking vendors for a comparable quote.