How to hire for potential: telling learning ability from enthusiasm
Some roles are better filled by someone who can learn the job than by someone who already knows it. When that is true, and when it is not, is in skills or fit: which to hire for and which to train. This post is about the harder part: once you have decided to hire for potential, how do you tell who has it? Every candidate says they are a fast learner. Very few interviews check.
What potential looks like, in practice
Potential is not a personality type. It is a short list of things a person does when faced with work they have not done before. They ask questions that show they understood what they were told. They use new information straight away rather than falling back on what they knew. They notice when they are stuck and say so. They remember what went wrong last time and do it differently. And they want to do this work, not just any work, because the willingness to spend six months being the least competent person in the room is not a general trait; it depends on caring about the thing.
Each of those can be evidenced. None of them is the same as being keen, articulate or likeable, which is what most interviews end up measuring.
Ask for evidence of learning, not enthusiasm for it
The useful questions are about the last time the candidate had to learn something for real. What did they have to pick up? How did they go about it? How long did it take? What did they get wrong first? What would they do differently now? A candidate who has actually learned things has specific, slightly embarrassing answers. A candidate who has not gives you "I'm a quick learner and I'm always reading."
Look for the shape of the story rather than the subject. Someone who taught themselves to run a bar's stock system, or got a returning-to-work qualification while raising children, or went from infantry to logistics planning, has shown you the thing you are hiring for. The setting does not need to be insurance.
Test it live
The best evidence is watching someone learn in front of you. Take a small piece of the actual work, explain it for five minutes, then give the candidate twenty or thirty minutes with it. It does not matter whether they finish. What matters is how they approach something unfamiliar: whether they use what you just told them, whether their questions are good, whether they check their own work, what they do when they hit a wall.
Keep the exercise the same for every candidate for the role, score it against written criteria, and have two people score. A candidate who learns visibly in half an hour will learn on the job. One who performs polish and cannot be taught anything in the room has told you something as well.
Aptitude tests: a floor, not a ranking
For roles where you are hiring people who do not yet have the skills, numerical and logical reasoning tests are a reasonable predictor of how quickly someone will pick up analytical work such as pricing, underwriting support or claims assessment. Entry-level programmes and Recruit-Train-Deploy models use them for that reason.
Two cautions. Use the test as a floor that screens out people who would struggle, not as a league table that crowns the highest scorer, because above the floor the test tells you little about who will be good. And use the same test, at the same stage, for every candidate for the role, with the results kept, so the process is consistent and defensible. A test that measures something the job never asks for is a filter on the wrong thing, and it will filter out people you wanted.
Assessing potential fairly
Hiring for potential is often sold as a route to a more diverse workforce, because it stops requiring experience that some groups have had fewer chances to get. That only holds if the assessment of potential is itself structured. An unstructured "do they seem bright and keen" judgement is the most bias-prone assessment there is. Same questions, same exercise, same test, written criteria, two scorers. How to keep the whole process fair is in building an inclusive workplace begins with your hiring. Pre-employment checks still apply in full, and are covered in the hidden cost of slow pre-employment checks.
What has to be in place before the hire
Hiring for potential moves cost from the salary line to the development line. A business that hires the attitude and then leaves the person to find the skills for themselves gets the worst of both: a lower-paid hire who takes a year to be useful and leaves when they are.
Before the offer goes out, name the person who will train the new hire and take that time out of their targets. Write down what the hire should be able to do at four, eight and twelve weeks, so probation is a set of checkpoints rather than a feeling. Agree when they are expected to work unsupervised. Decide who checks their work until then. If the business cannot commit to any of that, buy the skills instead, or bring the training in from outside.
Where this is already done at scale
Our sister company FlarePeople runs this model for insurance pricing. It recruits STEM graduates, ex-military people and career returners on aptitude and attitude, trains them in a ten-week academy covering pricing tools, the insurance sector, business acumen and the working skills the job needs, and deploys them to insurers on an initial 24-month contract with the option to take them on permanently. The assessment, the training and the first two years' supervision are handled for the client. How it works is in Recruit-Train-Deploy for insurance pricing.
Where Vermelo comes in
When a client is hiring for potential, we build the assessment into the process: the learning questions, the live exercise and, where the role warrants it, the aptitude test, scored the same way for every candidate. That is part of how our embedded talent teams run a client's hiring. If you are not sure which model fits the roles you have, the talent solutions finder is a two-minute start, or talk to us.
Assessing potential, answered
Ask for evidence of learning, not enthusiasm for it. Everyone says they love to learn. The useful questions are about the last time they had to: what they had to pick up, how they went about it, how long it took, what they got wrong first and what they would do differently. Then test it live. Give the candidate a short piece of the actual work with a few minutes of explanation, and watch how they approach something unfamiliar: do they ask good questions, do they use what you just told them, do they notice when they are stuck. A candidate who learns visibly in a forty-minute exercise will learn on the job. One who performs polish and cannot be taught anything in the room is telling you something too.
For roles where you are hiring people without the skills yet, yes, provided the test measures something the job uses. Numerical and logical reasoning tests are a reasonable predictor of how quickly someone will pick up analytical work such as pricing, underwriting support or claims assessment, which is why entry-level and Recruit-Train-Deploy programmes use them. Two cautions. Use the test as a floor, not a ranking, so it screens out people who would struggle rather than crowning the highest scorer. And use the same test, at the same stage, for every candidate for the role, with the results kept, so the process is defensible. A test that measures something the job never asks for is a filter on the wrong thing.
Someone to teach them, time to learn, and a way to tell whether it is working. Hiring for potential moves cost from the salary line to the development line, and businesses that hire the attitude then leave the person to find the skills for themselves get the worst of both. Before the hire, name the person who will train them and take that time out of their targets. Write down what the new hire should be able to do at four, eight and twelve weeks, so probation is a set of checkpoints rather than a feeling. Agree the point at which they are expected to work unsupervised. If the business cannot commit any of that, buy the skills instead, or bring the training in from outside: that is what a Recruit-Train-Deploy model does for you.