Back to blogs
Data & Research

We Analysed 5,000 Resumes. Here's What Actually Predicts a Good Hire.

Instinct-driven resume screening misses the strongest candidates. Our analysis of 5,000 resumes across 12 industries reveals what actually predicts hiring success — and what does not.

TwynIt Research · TwynIt Team·8 min read·Apr 28, 2026

Key metrics

5,000+ resumes analysed across 12 industries
Skill coverage is 3× more predictive than job title or company name
Employment gaps show no significant correlation with performance

What We Set Out to Understand

Resume screening is one of the most judgment-dependent parts of hiring. Recruiters build instincts over time — a sense of what a strong resume looks like. We wanted to test whether those instincts hold up against data.

We looked at 5,000 resumes across 12 industries, comparing screening decisions against actual outcomes: did shortlisted candidates make it to interview? Did they receive offers? How did they perform in the first 90 days?

Finding 1: Skill Coverage Is the Strongest Predictor

The single most predictive factor for shortlisting success — and subsequent offer — was skill coverage: how many of the JD's required skills appeared in the candidate's profile.

  • Candidates with high skill coverage but unrecognised company names: shortlisted 68% of the time
  • Candidates from well-known companies with low skill coverage: shortlisted 34% of the time
  • Performance difference between these two groups in the role: minimal

Finding 2: Resume Format Is Nearly Irrelevant

Teams frequently deprioritise resumes for formatting reasons — inconsistent bullet points, long paragraphs, non-standard layouts. Our data found that formatting explained less than 5% of the variance in shortlisting outcomes when skills and experience were controlled for.

Candidates filtered out for formatting reasons are often rejected for something that does not predict job performance. AI scoring removes this entirely — it evaluates substance, not presentation style.

Finding 3: Employment Gaps Are Misused as a Filter

Employment gaps were flagged as negatives in manual screening at a high rate. Our analysis found no statistically significant relationship between gaps and interview performance or 90-day outcomes.

  • Gaps under 3 months: no measurable impact on role performance
  • Gaps of 3–12 months: slight positive correlation in technical roles (period of directed learning)
  • Gaps over 12 months: worth exploring in conversation, but not a screening-stage filter

Finding 4: Notice Period Is Not a Proxy for Candidate Quality

Immediate availability and 60-day-notice candidates performed identically in role outcomes. Yet notice period was used as a de facto filter in 63% of the manual shortlisting decisions we reviewed.

Filtering by notice period at the screening stage likely eliminates strong candidates who simply have longer commitments, not less desirable profiles.

What This Changes About How You Screen

  • Build your shortlist on skill coverage and experience fit — not company pedigree or formatting
  • Stop using notice period as a screening-stage filter
  • Do not penalise resume formatting unless the role explicitly requires documentation quality
  • Employment gaps should prompt a conversation in the interview, not an automatic rejection

The Takeaway

The resume is a rough signal. AI scoring sharpens it by isolating what actually predicts good candidates: skills, experience depth, and relevance to your specific role. The proxies that shape manual review — brand recognition, formatting preferences, gap anxiety — are measurable, removable, and largely uncorrelated with the hire quality you are trying to predict.

Next step

Use data-driven scoring on your next hire. Start free on TwynIt — no credit card required.

Found this useful? Share it.

Your network might be facing the same hiring challenges. Pass it on.