An HRIS purchase reaches a CEO or a board as a tooling decision with a licence cost attached. It is not one. It is a decision about where your employment, payroll and compensation data will live for the next five to seven years, how hard it will be to leave, and what proportion of your HR team's time will be spent operating software rather than doing HR.
The questions below are written for the person approving the spend rather than the person running the evaluation. They are deliberately uncomfortable — a vendor and an internal sponsor who have spent three months on a shortlist will have rehearsed the feature comparison and will not have rehearsed these.
Group one: the business case
1. What is the problem, stated without naming a product category?
If the answer cannot be given without the words "HRIS" or "platform," the problem has not been diagnosed. Good answer: "Payroll takes eleven working days because attendance and salary revisions live in different places, and we have made corrections after disbursement in four of the last six months." Bad answer: "Our current system is outdated."
2. What does the current state cost, in money and hours?
Licence spend, reconciliation hours, correction volume, integration upkeep. If nobody has counted, the business case is a preference rather than an analysis. This is what our stack TCO calculator exists to structure.
3. What happens if we do nothing for twelve months?
A real answer names a specific consequence with a date attached — a compliance deadline, a headcount threshold, a system reaching end of support. Bad answer: a general statement about falling behind.
4. Who owns the outcome, and what will they be measured on?
Not who runs the project — who is accountable twelve months after go-live for the metric in question 2. Software purchases without a named owner and a named metric reliably become software purchases without a measured result.
Group two: the data
5. Where does each fact live today, and how many copies will exist after?
The strongest predictor of whether an HR system reduces work is whether it reduces the number of places the same fact is stored. A system that adds a sixth place where salary is recorded will add reconciliation, however good its interface. Ask for the after-state diagram, not the before-state one.
6. Which integrations are we relying on, and who maintains them?
For each connection: who built it, who fixes it when an API changes, what the response time commitment is, and whether it is covered by the licence or billed. "It integrates with everything" is a marketing claim; ask which of those integrations the vendor maintains versus certifies versus has merely seen someone build once.
7. What happens to payroll history, statutory records and part-year tax positions during migration?
This is where migrations fail. Historical payslips, year-to-date tax, leave balances, gratuity service dates and vesting schedules all have to arrive correctly, and their correctness is not obvious until a year-end or an audit. Ask specifically who validates the migrated data and against what.
8. Can we get all our data out, in what format, and how quickly?
Ask for the export specification during evaluation, not at renewal. A good answer describes complete structured exports including historical payroll and audit trails, available on demand, in a documented format. A bad answer offers reports rather than data, or requires a support ticket and a professional-services quote. The moment to negotiate exit terms is while the vendor still wants your signature.
Group three: the commitment
9. What is the total cost over five years, including everything that is not licence?
Implementation, migration, integration, training, the parallel-run period, annual uplift, and the modules quoted separately. Ask for the uplift cap in writing; an uncapped annual increase on a system you cannot easily leave is a structural weakness, not a pricing detail.
10. What are we giving up?
Every consolidation trades depth for coherence, and every best-of-breed choice trades coherence for depth. If the sponsor cannot name what gets worse, they have not evaluated honestly — and the team that loses its specialist tool will discover the trade-off after go-live rather than before.
11. What is the concentration risk, and what is our position if the vendor fails or is acquired?
Consolidating hiring, HR, payroll and equity data into one vendor is a commercial dependency. Ask about financial stability, ownership, and what contractual protections exist on acquisition. This is a reasonable question for any vendor of any size and a revealing one.
12. Who else at our size and in our jurisdictions is live on this, and may we speak to them unaccompanied?
The qualifier does the work. Reference calls arranged and attended by the vendor produce reference-call answers. Ask specifically for a customer in your regulatory environment — a payroll system that is excellent in one country can be a thin wrapper in another.
Three answers that should stop a purchase
"We'll work out the migration during implementation." Migration scope is the largest source of cost overrun in this category. If it is unscoped at approval, the number in front of you is not the number.
"The integrations are on the roadmap." Roadmap means it does not exist. If the business case depends on it, the business case depends on a future decision by someone who does not work for you.
"Data export is available through our professional services team." That is a statement about switching costs disguised as a service offering. Resolve it before signing or accept that you are choosing a system you cannot leave.
A note on what these questions are not
None of the twelve is a feature question, and that is deliberate. Feature evaluation is the part your team is already doing well and the part vendors are organised to answer. The approver adds value by testing the things a demo cannot show: what the data architecture will look like afterwards, what the commitment really costs, and how hard it will be to reverse.
If the answers to questions 1, 2 and 8 are strong, most of the rest tends to follow. If those three are weak, no feature comparison will rescue the decision.
Frequently asked questions
Should a CEO be involved in an HRIS selection at all?
In selection, no. In approval, yes — the switching cost and data-ownership consequences outlast most executive tenures and are not visible in a feature matrix.
How long should an evaluation take?
Long enough to get written answers to questions 7, 8 and 9. If a vendor's commercial timeline does not allow that, the timeline is a negotiating tactic rather than a constraint.
Is a single suite always better than best-of-breed?
No. Consolidation reduces reconciliation and increases dependency; best-of-breed does the reverse. The right answer depends on whether your pain is coherence or capability — and question 10 is designed to force that discussion into the open.
What if we already bought and these questions were not asked?
Ask 8 and 9 now anyway. Knowing your export position and your five-year cost is valuable at any point, and both are easier to establish while a relationship is healthy.
This is a decision framework rather than a data piece and contains no benchmark claims. Helion sells HR and payroll software and therefore competes in the category this article discusses. General information, not procurement or legal advice; contractual terms including data export and uplift provisions should be reviewed by your own counsel.