A skills audit tells you what your people can do now, measured against what the work needs. Done well, it stops you hiring for skills you already have and training people in things they already know.
Why bother
Most companies know job titles, not skills. A title says what someone was hired to do, often years ago. It does not say that your finance analyst writes the best SQL in the building, or that three team leads have never had a hard conversation with anyone.
An audit turns that guesswork into a map. You can then answer real questions: who could step up if a manager left, which team is one resignation away from trouble, and where training money will make a difference.
The map also goes out of date faster than it used to. The World Economic Forum expects that 39% of workers’ existing skills will be transformed or become outdated between 2025 and 2030.
1. Pick the question before the spreadsheet
Start with the decision the audit has to support. “Do we have the skills to open a second site next year?” is a good question. “Let’s see what everyone can do” is not, because it has no end.
The question tells you who to include. A first audit of one department or one job family is often worth more than a company-wide one that never finishes.
2. Keep the skills list short
For each role in scope, list the skills that matter for doing it well. Aim for eight to twelve per role. Fifty skills per person produces fatigue, not data.
- Mix technical and human skills. “Forecasting” and “giving feedback” both belong.
- Use the words people use. If the team says “reporting”, do not call it “data storytelling”.
- Share skills across roles where you can. One “Negotiation” skill used by Sales and Procurement lets you compare the two.
3. Agree what each level means
A rating is only useful if two people mean the same thing by it. Pick a simple scale, four or five levels, and describe each one in terms of behaviour you can see.
Avoid labels like “expert” on their own. Everyone reads them differently, and most people are too modest or too generous with them.
4. Rate from more than one side
Self-ratings alone are unreliable. People who are new to a skill tend to rate themselves high; people who are good at it tend to rate themselves low. Combine views:
- The person rates themselves first, so the manager does not anchor them.
- The manager rates the same skills, and notes an example for any big difference.
- A peer or two rate the skills they see every day.
- Evidence backs it where it exists: a certificate, a finished project, a course passed.
Tell people what the ratings are for, and what they are not for. If they fear the audit feeds straight into pay or redundancy, you will get careful answers, not true ones.
5. Read the gaps
Compare each person’s ratings with the level their role needs. Then step back and look for patterns, not individuals:
- Team-wide gaps point to training, or to a hire.
- Single points of failure: a skill only one person has at the level needed.
- Hidden strength: people rated well above their role, who may be ready for more.
- Big rating disagreements between a person and their manager, which are worth a conversation on their own.
6. Act, then repeat
Share the results with everyone who took part, at least their own. Then pick three actions and name an owner for each: a course for a team-wide gap, a stretch project for someone with hidden strength, a second person trained on a single-point skill.
Skills change, so the audit is never finished. Re-rate every six or twelve months, and when someone changes role. The second audit is where the value shows, because now you can see movement.