The rise of ‘skills-based’ hiring has transformed the way talent teams screen prospective candidates.
Assessments of technical skills and human capabilities are now the signals of job performance that hiring managers trust the most, outranking interviews and CV screens by humans and AI.
Despite this pressing need, many are unable to effectively run such assessments. This causes several problems:
- Slow recruitment processes and frustrated team managers
- Ill-informed decisions on candidates
- Costly mis-hires, particularly for candidates employed across borders
So, how can talent teams build more rigorous skills assessments they can trust?
Why effective skills assessments matter
The shift to ‘skills-based’ hiring started with talent teams who felt traditional credentials failed to give a fair or accurate indication of the strength of a candidate.
As Emma McDowell, Head of Client Delivery at Flex Legal, a recruitment firm for the legal sector explains: “Skills assessments allow employers to look beyond CVs, grades, and academic background.”
Instead, hirers wished to conduct their own evaluation of whether candidates could demonstrate the necessary skills for the role.
This has been compounded by the devaluation of the CV or résumé, and the overwhelming rise in volume of applications received for each vacancy.
According to Greenhouse’s 2026 benchmark report:
- Annual applications per organization have increased 128.1% since 2022
- Recruiters per organization have decreased 55.6% since 2022
- Annual applications per recruiter have increased 411.8% since 2022
Moreover, organizations have lost trust in what candidates claim. 2026 Arctic Shores research finds that while 87% of candidates say their CV represents them accurately, only 51% of hiring managers agree. The reverse also applies, with candidates also losing faith in hiring processes.
At the same time, jobs are becoming less static; employers recognize that work can be split into tasks, with employees able to work flexibly, rather than being confined to a traditional role.
Skills-based hiring has therefore the default approach for businesses, in theory. But in practice, most struggle to effectively assess skills.
The first hurdle: Defining skills
To begin, many employers fail to establish or agree on the actual skills they are recruiting for.
It may seem obvious, but particularly for large, global companies, defining what skills should be prioritized is no mean feat — just ask anyone who’s tried to build a company-wide skills taxonomy.
Creating shared agreement with the hiring manager and other stakeholders on the skills that need to be screened, and how they will be measured, is vital.
“Consider what skills are actually required for the role, otherwise suitable candidates will slip through the net,” warns McDowell.
Christy Belsky is Recruiting Programs & Operations Manager at Huntress, a cybersecurity company with more than 500 employees around the globe.
She agrees that building an agreement on skills is the first step recruiting program owners must take.
Her advice is to begin with an interview calibration meeting at the beginning of the hiring process.
“Interviewers should leave the calibration meeting with a clear understanding of the skills they are responsible for assessing and the questions and tools that they will use during the interview,” Belsky says.
Align on assessment methodology
Alignment on a consistent assessment methodology itself is the next step.
Belsky argues that enabling interviews with the right tools and questions to accurately assess specific role requirements is the “largest challenge.”
“A structured interview process, including an interview plan, assessment questions and evaluation criteria should be established and applied consistently,” she explains, noting that best practices depend on the skill being evaluated.
Technical skill sets, like coding, may mean using a dedicated testing platform. These can be configured to “address the specific skills required to do the job and apply a consistent standard across your candidate slate,” Belsky says.
Alternatively, the recruiting manager suggests employers can ask for “scenario-specific examples from the candidate’s professional experience,” or run other assignments that demonstrate their skill level.
McDowell adds that tasks set should be “tasks relevant to the role, rather than generic ones; ensure assessors are trained; and have a structured scoring system.”
“Skills like collaboration, leadership or adaptability are hard to assess, so consider situational judgement tests or structured exercises,” she offers.
Whatever the methodology, recruiting teams should be careful to support other stakeholders with the implementation of the assessment in practice.
Belsky advises: “Hiring managers and other interviewers require regular coaching and access to resources to maintain a consistent, relevant skill assessment practice that ensures protocols are understood and applied at scale across the organization.”
Communicate & ensure consistency
David Kemmerer, co-founder and CEO of crypto tax software provider CoinLedger, has built a small but global team. He goes a step further than McDowell and Belsky, proposing a method to guarantee alignment with the skills needed for the role.
Rather than setting an assessment based on a hypothetical scenario, he tells The Exchange that employers should instead “add the candidate to any part of a recently completed project, pay them for consulting and see what they produce.”
“How do they deal with messy situations that they’ve not dealt with before? Do they ask for help or not?” he asks. “It tests how they perform under pressure instead of continuously asking questions they’ve rehearsed answers to.”
Whether the assessment is hypothetical (but relevant), or real, the key is to build firm guidelines around any assessment and communicate them clearly — not just with hiring managers or interviews, but also with candidates.
Otherwise, candidates will not be able to complete the test on an even footing. An unfair advantage for one, or confusion for another, may lead to results that cannot be trusted.
“Communicate what the hiring process will look like,” McDowell sets out. “Set clear rules on AI usage; allow candidates to request feedback; proactively offer reasonable adjustments to candidates, and ensure consistency.”
‘We lost out on an incredible senior engineer’
There are also specific difficulties when running assessments during cross-border hires, including technical and cultural friction points. Among other problems, these can lead to firms losing out on top candidates.
Again, consistent conditions and communications for each candidate are vital.
“The main challenge is infrastructural inequality,” Kemmerer says, speaking from painful experience. “Last year, we lost out on an incredible senior engineer in a different country because our standard async test required downloading a massive 4GB container environment.”
The candidate’s poor local internet bandwidth turned a two-hour test into a six-hour wait and they ultimately were unable to complete the process.
Kemmerer has also encountered what he calls “psychological friction.”
“In Western tech culture, ‘fail fast and iterate’ works. In some countries, candidates view an imperfect test submission as an absolute failure,” the founder explains. “They spend triple the allotted time polishing a minor detail because they are terrified that sending something incomplete will reflect poorly on their character.”
“A candidate in another region will ship a brilliant draft with missing links within the time frame,” he adds. “It is hard not to judge the candidates by the same measure.”
This where firm and clearly-communicated assessment criteria comes into play. CoinLedger has resolved these challenges by setting up a GitHub to ensure every candidate can access the same infrastructure; and improving its guidelines to account for cultural differences and establishes clear rules for the task.
“We rewrote our brief to emphasize that we reward prioritization over perfection,” Kemmerer says. “ There is a bolded line that says, ‘Submit what you have once the timer runs out. If you think it is broken, write a three-sentence note on how you would fix it if you had more time’.”
