A polished answer is not proof someone can do the job
Work simulations can make candidate capability more visible—if you design them around the right task.
How to design a work sample assessment that shows what a candidate can actually do.
“Tell me about a time you solved a difficult problem.”
A candidate gives a convincing answer. They explain the situation, describe their actions and finish with an impressive result.
That is useful information. But you still have to establish what they personally contributed, how difficult the problem was, and whether they could apply that judgement in your environment.
A short work simulation gives you another piece of evidence.
Let candidates demonstrate the work
Imagine hiring an operations analyst. Give candidates a fictional case containing conflicting records, a customer complaint and an approaching deadline.
Ask them to explain:
- What would they investigate first?
- Which information is missing?
- What could they resolve themselves?
- When would they escalate?
- How would they communicate the next step?
This makes their approach available for examination. You can see whether they notice inconsistencies, make unsupported assumptions or recognise an important risk.
Work samples are an established assessment method. The US Office of Personnel Management describes them as tasks that closely resemble work performed in the role, while noting their development and administration costs. See its guidance on work samples and simulations.
Design the scoring before reviewing candidates
Decide what a strong response looks like before anyone completes the exercise. Give candidates consistent instructions, explain whether AI tools are allowed, and provide appropriate adjustments.
Keep the task proportionate. Use fictional material and avoid asking candidates to produce unpaid deliverables for your business.
Also distinguish between skills needed on entry and knowledge you plan to teach. OPM cautions that work samples may be inappropriate when the tested activities will be taught after selection.
Keep the conclusions proportionate too
One exercise provides evidence about performance on that exercise. It cannot establish everything about a person’s future performance.
An AI-generated answer on a candidate’s behalf is not evidence that the candidate can do the work.
Pontiq’s proposed Evidence Passport is intended to keep those distinctions visible: what the candidate demonstrated, under what conditions, and what still needs to be explored.
Hiring for a role that stays open too long? A Pontiq discovery engagement takes one role and works out what it takes, where the people are, and how to tell. See how the method works.
Discuss a discovery engagementTalent leader with a view on this? Take part in our research conversations.