Beyond Keywords: Recognizing Soft Skills in Hiring

The keyword problem in recruitment

'Team player', 'excellent communicator', 'proactive'. Every CV contains them. Every job description asks for them. But what do they actually mean? And how do you assess whether someone truly has them?

The problem with keywords is that they say nothing. Everyone calls themselves a team player. Nobody writes on their CV: 'I prefer working alone and communicate poorly.' Keywords are noise, not signal.

Yet soft skills are often the decisive factor in successful hires. Technical skills can be trained. But empathy, adaptability, and collaboration? Those run deeper. And they determine whether someone can not only perform the role, but also thrive in the team.

This article shows how to assess soft skills in a way that goes beyond keywords and surface impressions.

Why soft skills are underestimated

The recruitment industry has focused on hard skills for years. Years of experience, degrees, technical certifications. These are easy to measure, easy to compare, and easy to fill in on an ATS form.

Soft skills are harder. They're contextual (an excellent communicator in one organization might struggle in another), they're difficult to test in a short conversation, and they're subjective (what I call 'assertive,' someone else calls 'aggressive').

But the research is clear: soft skills are at least as important as hard skills for long-term success. A study by Harvard, Stanford, and the Carnegie Foundation concluded that 85% of job success depends on soft skills, and only 15% on technical knowledge.

The five soft skills that matter most

1. Communication

Not just 'being a good talker.' Communication means listening, choosing the right tone, conveying complex information clearly, and adapting to your audience. The difference between a good and a great employee often comes down to communication.

How to spot it in an interview? Watch whether the candidate actively listens (summarizes what you said), asks questions before answering, and can articulate complex thinking clearly.

2. Adaptability

The world is changing fast. The employee who succeeds in 2025 isn't the one who knows the most, but the one who learns fastest and adapts most easily to new situations.

How to spot it? Ask about situations where plans changed. How did the candidate react? Was there frustration, acceptance, or enthusiasm? People who handle change well see it as opportunity, not threat.

3. Collaboration

Collaboration is more than 'being nice to your colleagues.' It's about sacrificing personal interest for team results, actively contributing to others' success, and dealing constructively with conflicts.

How to spot it? Notice how the candidate talks about previous teams. Do they use 'I' or 'we'? Do they give credit to others? Do they describe conflicts as learning moments or as someone else's fault?

4. Problem-solving ability

Not every problem has an obvious solution. The employee who can analyze problems, weigh options, and make a thoughtful decision is worth their weight in gold.

How to spot it? Give the candidate a hypothetical problem and focus on the thinking process, not the answer. Does the candidate ask clarifying questions? Consider multiple options? Or jump straight to a solution?

5. Emotional intelligence

The ability to recognize and regulate your own emotions, and to understand and respond to others' emotions. In teams, this is the lubricant that keeps everything running smoothly.

How to spot it? Look for self-awareness. Can the candidate name their own weaknesses without getting defensive? Do they respond empathetically to others' feelings? How do they handle stress or disappointment?

The limitations of traditional assessment

Traditional interview methods are poor at assessing soft skills. Why?

• Candidates give socially desirable answers ('yes, I'm absolutely a team player')
• Interviewers have their own biases (extrovert = good communicator? Not necessarily)
• Short conversations give an incomplete picture of someone's behavioral patterns
• Without structure, every interviewer assesses different things

The result: soft skills are assessed based on impression, not evidence.

A better approach: observe behavior, don't weigh words

The key to assessing soft skills is behavioral observation. Not what the candidate says they would do, but what the candidate actually does.

Method 1: Behavioral interview questions

'Tell me about a situation where you disagreed with a team member about the approach.' This question forces the candidate to describe concrete behavior. Look for details: who said what, how did it end, what did the candidate learn?

Generic answers ('I would discuss it and find a solution') are red flags. Specific answers with context, action, and result are green signals.

Method 2: Observation during the interview

The interview itself is a test of soft skills. How does the candidate communicate? Do they listen? Do they ask questions? How do they handle an unexpected question or a silence?

Insights analyze communication patterns automatically. What's the talk/listen ratio? How deep are the questions? How does someone react to unexpected turns? This data supplements your subjective observation with objective metrics.

Method 3: Work-related assignments

Give the candidate an assignment that draws on the soft skills you're looking for. A group assignment for collaboration. A presentation for communication. A case study for problem-solving ability.

Method 4: References targeted at soft skills

Many reference checks are superficial. 'How was it working with John?' Instead, ask targeted questions: 'Can you give an example of how John dealt with a conflict in the team?' or 'How did John react to unexpected changes?'

How AI helps with soft skills assessment

AI can't measure soft skills. But AI can recognize patterns that are relevant to the assessment.

AI summaries automatically flag moments in the conversation relevant to competency assessment. A moment of empathy, an example of problem-solving, a reaction to pushback.

Transparency links every observation to the exact moment. So you can discuss the same moments as a team and assess them together, rather than everyone relying on their own impression.

And with data extraction, relevant observations are automatically captured in your CRM, ready for comparison with other candidates.

The scorecard model for soft skills

To assess soft skills consistently, you need a scorecard. Here's a workable model:

For each soft skill relevant to the role:

• Score 1 (Insufficient): No evidence of the skill. Counter-evidence present.
• Score 2 (Limited): Some evidence, but inconsistent or superficial.
• Score 3 (Satisfactory): Consistent evidence of the skill in relevant context.
• Score 4 (Strong): Convincing evidence, multiple examples, clear reflection.
• Score 5 (Exceptional): Inspiring evidence that exceeds expectations.

Record scores via CRM data entry immediately after the conversation. The faster you score, the more reliable your assessment.

The cultural fit trap

'Does this person fit our culture?' is one of the most dangerous questions in recruitment. Not because culture fit isn't important, but because it's often an excuse for bias.

'Doesn't fit the culture' can mean: 'thinks differently than us,' 'comes from a different background,' or 'doesn't match my image of what a colleague looks like.' That's not cultural fit. That's discrimination.

Better is 'values alignment': does the candidate align with the organization's values? That's concrete and testable. You can ask about how someone handles transparency, feedback, accountability, and collaboration. Those are values, not vague 'fit.'

Human-AI collaboration in soft skills assessment

AI is not a replacement for human judgment in soft skills assessment. It's a complement. The recruiter conducts the conversation, reads body language, and feels the energy in the room. AI analyzes the content: what words does the candidate use? How detailed are the answers? Does the language match the role? The combination of human intuition and AI analysis produces a more reliable picture than either alone.

In practice, this works best when recruiters compare their own observations with the AI analysis. Do they reach the same conclusion? If not, that's a valuable discussion point. Perhaps the recruiter picked up on something the AI missed, or vice versa. This dialogue between human and technology makes the assessment stronger than either judgment alone.