Five checks that separate real skill from a good résumé

Hiring someone you will never meet in person removes most of the signals you normally rely on. No shared office, no mutual contacts, no sense of how they work. What is left is a document the candidate wrote about themselves, and a conversation they have had practice at.

That is not an argument that candidates lie. It is an argument that a résumé measures presentation skill, and presentation skill is only the job if you are hiring for presentation. The five checks below each replace a claim with evidence.

A small, scoped task that mirrors the real job beats any interview claim. Not a puzzle, not an algorithm quiz — a slice of what they would actually do in week one.

Keep it short enough that a good candidate will do it unpaid without resentment. If the task takes more than about an hour, you are no longer screening, you are asking for free work, and the strongest candidates will decline first.

Not "tell me about a project" — ask for a specific decision and the trade-offs they weighed. What were the options, what did they give up, what would have changed their mind.

Depth shows fast here. Someone who did the work remembers the constraint that made it hard. Someone who watched it happen remembers the outcome.

Real practitioners have opinions about their own past work, usually quite specific ones. Fakers deflect, or answer with something generic about communication.

This question is also the cheapest seniority signal available. Juniors tend to say they would have worked harder. Seniors tend to say they would have scoped it differently.

Take a headline claim and dig one level below it. If someone says they led a migration, ask what broke. If they say they owned a metric, ask what it was before and after.

The specifics either match the headline or they do not. You are not trying to catch anyone out — most inconsistencies are honest compression. But the ones that matter show up here rather than in the résumé.

Decide what good looks like before you see the answers. Strengths, gaps, and a confidence signal per answer — written down, so that two candidates assessed a week apart are assessed the same way.

This is the check most teams skip, and it is the one that does the most work. A rubric is what turns "I liked them" into a reason you can give a colleague, defend to a candidate who asks, and compare against the next person.

Nothing above needs software. The reason teams skip these checks is not that they disagree with them — it is that writing a job-specific task and a rubric for every role takes an afternoon, and the fifth candidate arrives before the first rubric is finished.

That time cost is the actual problem. It is also the only part of this a tool can solve.