Shortlisting is ranking, not reading. Here's a process that turns a stack of CVs into a defensible top list.
Shortlisting goes wrong when it's done as sequential reading — by CV fifty, your bar has drifted and your memory of CV three is gone. Treat shortlisting as a ranking exercise against fixed criteria and it becomes fast and defensible.
1. Fix the criteria before you start
Pull the must-haves from the job description. These are the axes every candidate is scored on — no new criteria invented mid-pile.
2. Score every candidate against the role
Give each CV a fit score against those criteria rather than a gut yes/no. Scoring the whole batch the same way is what makes the ranking trustworthy. An AI screen does this in parallel across up to 100 CVs.
3. Rank, then read the top
Sort by fit and read the top of the list closely, confirming the score against the evidence. You spend your attention where it matters instead of reading every CV in full.
4. Document why
Record the reason each candidate made or missed the shortlist. It speeds up calibration with the hiring manager and protects you if a decision is ever questioned.
About CV Screentool: AI-powered CV screening for recruiters and hiring teams. Score candidates against any job description, rank applicants, surface skills gaps, and generate tailored interview questions in minutes.