How to design a role-based assessment step by step

You have the role open, the drive to fill it well, and a folder of possible assessments you can’t quite put in order. Should I send a logic test? Add a personality one? Ask for a case study? The problem is almost never a lack of tools: it’s not knowing what to combine for this specific role, without ending up with a 90-minute assessment nobody wants to take.

The good news is that designing an assessment by role isn’t guesswork. It’s a short, orderly path: you start from the real pain of the position, translate it into competencies, choose which test measures each one, and finally balance how much you ask based on the experience you expect. Let’s walk that path.

Start with the role’s pain, not the tests

Before looking at the catalog, answer an uncomfortable question: what happens when someone does this job badly? An analyst who makes a mistake in a reconciliation, a support agent who loses their cool with a difficult customer, a salesperson who promises what they can’t deliver. That “what goes wrong” is your best compass.

Talk with the person who’ll be the direct manager and with someone who already does that role well. Ask about the moments that separate a good fit from a mediocre one: the real situations, not the ones in the job description. That’s where the competencies that truly matter come from —not the ones that sound good in a posting.

Translate the pain into 3 or 4 key competencies

After those conversations you’ll have a long list. Trim it. A useful assessment measures a few competencies well, not many halfway. For most roles, three or four are enough: one technical to the trade, one cognitive (how they reason and solve), and one or two behavioral tied to how your team works.

Be specific about the LATAM context: a “communicator” isn’t the same for someone serving customers over chat in Bogotá as for someone negotiating in person in Lima. Name the competency with the role in mind, not in the abstract.

Choose the test that measures each competency

This is where most people get stuck, and it’s simpler than it looks: each competency is paired with the type of evidence that shows it best. Numerical reasoning for roles with data, situational judgment for roles dealing with people, a case study for the technical side of the trade. The rule is that each test exists for a reason, not because “it’s always included.”

If you don’t want to build this from scratch, reviewing combinations already thought out by role saves you hours. In the roles library you can see which competencies tend to carry weight in each position, and pages like financial analyst show how a concrete profile translates into a comparable signal across candidates.

Balance coverage and candidate experience

An assessment that covers everything but takes an hour and a half gets abandoned halfway, and you screen for patience, not talent. The trade-off is real: every test you add improves coverage and, at the same time, asks more time and energy of the person.

A practical rule: prioritize the two competencies that most distinguish performance and give them the main weight. The rest can be measured more lightly. Think too about the passive candidate, the one who has a job today and will only give you 30 minutes: if your assessment respects their time, you reach better profiles. To go deeper on how to dose this without losing signal, this guide on combining competencies by role gets into the detail.

Close with integrity controls and common criteria

A well-designed assessment also looks after the evidence being reliable, with integrity controls that give context to the results without turning the process into a witch hunt. The goal isn’t to distrust everyone: it’s that when you compare candidates, you do it on the same basis.

That’s what common criteria give you: instead of “I had a good feeling,” the team discusses comparable signals and reports to prepare the interview. Kokoro supports that decision by giving you evidence beyond the resume before you interview, but the team keeps the final decision. If you want to see how that end-to-end flow is built, take a look at the product.

In short

Designing an assessment by role is a four-step path: start from the concrete pain of the position, translate it into three or four competencies that truly distinguish, pair each one with the test that measures it, and balance coverage with the candidate’s time. Close with integrity controls so everyone compares on the same basis. Done this way, you stop guessing which tests to combine and start deciding, with backing, who to interview.