Run a representative trial and cost model
Course overview · 4 min reading + 12 min practice, estimated
Principles and method
Use the same realistic fictional scenarios across shortlisted options. Include ordinary tasks, edge cases and operator recovery. Record preparation, training, review and administration effort alongside licence costs. Consider migration, integrations, support, usage charges and exit costs over a stated horizon. Avoid extrapolating savings from a supplier-led best-case demo. Give operators time to use the product and collect evidence against the requirements. Keep the trial decision separate from enthusiasm about one impressive feature. A failed essential requirement should remain visible in the comparison.
Worked example
A cheap tool requires substantial manual cleanup and lacks usable export. A higher licence price may still have a lower total operating cost, but the comparison states assumptions rather than inventing savings.
Put it into practice
Build a one-year cost table for two fictional suppliers and a no-purchase option.
Use fictional information and keep your work in your own notes.
Compare your approach: self-review guidance
Show licence, implementation, labour and exit assumptions. Use ranges for uncertain costs. Include trial evidence and do not let a weighted total hide a failed mandatory control.
Sources and further reading
Original Academy teaching and fictional examples. These references provide context, not endorsement. Edition 2026.09; updated 2026-09-24.
- GOV.UK: Responsible AI in recruitment
UK guidance on procuring and deploying recruitment AI.
- NIST: AI Risk Management Framework
Voluntary framework for organising AI risks and controls.