Three Teams Asked for Staff Software, and Meant Three Different Things

Evaluations stall for a reason that has nothing to do with the products. Three people inside the same organisation asked for “a staff system”, each meant something different, and nobody said so out loud. Six weeks later the shortlist contains products that are not comparable, and the meeting where that becomes obvious is the one where the project loses momentum.

The three requests are easy to tell apart once you know they are three.

HR asked for the record

HR’s daily problem is that facts about people live in too many places. The contract is in one folder, the visa expiry in a spreadsheet, the last appraisal in an inbox, the leave balance in a second spreadsheet that disagrees with the first. What they want is a single authoritative file per person, and workflows that keep it current without chasing.

That is an HRMS: organised around the employee. Its natural unit is a person and its natural question is “what is true about this person, and what is about to stop being true?” Its hardest features are lifecycle ones — joining, changing, leaving — and its strongest ones are usually documents and approvals.

Operations asked for tomorrow

The operations manager has a different problem and a shorter horizon. Someone called in sick, a site is short, and the question is who can cover it without breaching a rest rule or pushing someone into overtime nobody approved. Nothing in an employee file answers that.

That is workforce management: organised around the day. Its natural unit is a shift at a place, and its natural question is “is tomorrow covered, and what does it cost?” Its hardest features are rostering, coverage rules and multi-site visibility — and in the UAE, specifically, sites that sit under different entities and staff who move between them. That last part is why the UAE version of this category is not simply the global one with a currency swapped.

Finance asked whether the hours are real

The third request is the quietest and the one that most often decides the purchase. Payroll is paying against numbers, and someone has to be able to say where those numbers came from — before the payment, not during a dispute afterwards.

That is an attendance and evidence system: organised around the event. Its natural unit is a check-in, and its natural question is “can this be shown to have happened, and if it was changed, by whom and why?” A correction history is not a nice-to-have here; it is the entire product. An attendance record with no visible corrections is not a clean record, it is an incomplete one.

Why the marketing does not help you

Every serious product lists all three categories, because the categories are what buyers search for. That makes a feature matrix nearly useless for telling them apart: everyone ticks everything, and depth is invisible in a checklist.

What separates them is which half was built first. A product with a deep employee record and shallow rostering is not deficient — it has a design centre, and so does the one with the opposite shape. The mistake is not choosing the wrong category. It is not knowing which one you are in, and therefore not knowing which weaknesses you have agreed to live with. If you want the market split up rather than the requirement, the category guide does that job.

The one-sentence test

Before anyone books a demo, write down the question this system will be asked most often, with the name and role of the person who will ask it. Not the full requirements list — one sentence.

• “Where is this person’s contract, and when does their visa expire?” — you are buying records.
• “Who is covering the second site tomorrow?” — you are buying operations.
• “Are these hours right, and can we show it?” — you are buying an evidence trail.

Most organisations need all three eventually. Almost none need all three equally at the same time, and the sentence tells you which one you are actually short of. Buy for that one, and require the other two to be adequate rather than excellent — that is a shortlist you can defend, and, more usefully, one you can finish.