How to run a skills audit that actually feeds your Workplace Skills Plan

Search for how to run a skills audit and you will get competent, well-written guidance about competency frameworks, skills ontologies and heat maps. Almost none of it is written for a country with a Workplace Skills Plan in it.

That matters more than it sounds. A skills audit run to international best practice can produce an excellent report that is useless for your submission, because it collected capability and skipped the fields your SETA requires. Then somebody rebuilds the data in April from payroll.

So this is the same exercise, shaped for what it has to feed.

What is a skills audit meant to produce?

Public sector guidance on skills audit methodology puts the purpose in one sentence. It is to identify the current skills and knowledge that an organisation has, and the skills and knowledge that it needs.

Both halves are load-bearing. An inventory of what you have, with no view of what you need, is a filing exercise. A wish list of what you need, with no baseline, is a budget request. The value is in the comparison, and the gap between the two columns is the thing the whole audit exists to surface.

The most common surprise is not a gap. It is capability nobody knew about. People arrive with skills from previous jobs, qualifications they never mentioned, study they did on their own, competences built on a different site years ago. None of it appears anywhere in the business until somebody asks, and an audit that surfaces two people who can already do the thing you were about to recruit for has paid for itself before the report is written.

There are two distinct versions worth naming, because they answer different questions. A personal skills audit assesses one individual's level against a specific or general skill and feeds a development plan. A group audit covers a team or the whole organisation and produces an inventory. Employers frequently want the second and commission the first, or the other way round.

Which decision makes or breaks it?

Scope, and it is decided before any data is collected.

The warning is blunt in the methodology guidance: avoid defining a broad scope, because there is no benefit in auditing every single skill in the organisation. Practitioner guidance says the same thing from experience, noting that too many audits are rushed, poorly executed and quickly out of date, precisely because they tried to cover everything.

The filter that works is business consequence. Audit the capabilities where being short actually costs you something: the statutory tickets, the trades, the roles that are hard to fill, the competences a tender asks about, the single points of failure where one person knows something nobody else does. Leave the long tail alone.

A narrow audit that gets finished and repeated beats a comprehensive one abandoned in month three.

What does the process look like?

Six steps, consistent across the better guidance, with the South African additions noted.

Define objectives and scope. Start from the business reason. Grant eligibility, scorecard position, employment equity targets, automation coming, a tender qualification requirement. The objective determines which fields you collect, and collecting the wrong ones is not recoverable later without asking everybody again.

Build the framework. A skills matrix or competency list setting out the capabilities each role requires and the proficiency levels against them. This has to exist before assessment begins, because an audit measures attainment against a framework and cannot invent one as it goes.

Collect the data. Job descriptions, qualifications, work experience and performance records, plus employee self-assessment and manager input. Other sources are already in the building: certificates, learning records, performance history, training registers.

Analyse against the framework. Where are the hidden strengths, where are you short, where is effort duplicated. This is the baseline.

Identify gaps and future needs. Look past current job descriptions to what the business will need in three to five years, then prioritise. A gap nobody is going to act on does not need to be in the report.

Act, then review. Translate findings into training, mentoring, redeployment, recognition of prior learning or recruitment, align them to the Workplace Skills Plan, and update the audit annually so it stays true.

How should you score it?

Simply, and consistently, because elaborate scales produce false precision from people guessing.

A four-point scale that works in practice runs from significant experience and expertise, through some knowledge or experience, to interested in learning more, to no knowledge or experience. The third point is unusual and useful, because it captures appetite rather than only capability, and appetite is what makes a development plan succeed.

Two cautions on self-assessment. Some groups consistently underestimate themselves and others are generous, so the scale needs explaining rather than just distributing. And self-scores need verification. A line manager review session to validate outcomes is a standard step for a reason, and it is the step that turns a survey into an audit.

What do South African employers have to collect that the guides omit?

This is the whole reason not to lift a template from elsewhere.

Your Workplace Skills Plan and Annual Training Report are built on a headcount broken down by occupational level, race, gender and disability status. Your B-BBEE skills element turns on who was trained, at what NQF level, and whether they were absorbed afterwards. Your employment equity plan works off the same demographic structure.

So a skills audit run here has to carry those fields alongside the competence data. For each person: occupational level, race, gender, disability status, qualifications held with NQF levels, and the training already received. Collect them once, in the audit, and three separate reporting obligations are fed from one dataset.

Skip them and you have a capability report plus a separate scramble through payroll every April. This is the single largest difference between an audit that saves work and one that adds it.

When is the right time to run one?

There is an annual rhythm and there are triggers, and the triggers are the ones most businesses ignore.

The organisational audit suits an annual cycle, ideally finishing in time to inform the plan rather than chasing it. Individual audits are event-driven, and the events are ordinary: a new role being created, somebody moving into a different role, a lateral move to another department, a project that needs particular skills, or a performance conversation where the question is what specifically this person needs to improve.

The underlying reason to keep at it is that the workforce does not hold still. Exits, new hires and changes in how work is done all move the picture, so a team is not the same as it was a year ago even if the headcount is identical. Surveyed organisations overwhelmingly report either having a skills gap now or expecting one within a few years, which is unsurprising, and the useful question is not whether you have one but where.

A restructure is the highest-value moment of all. When a process or technology change removes the need for a role, the audit answers the question that actually matters, which is what else the people in that role can already do.

How is this different from a training needs analysis?

Worth being precise, because the two get run as one exercise and both suffer.

The audit establishes what capability exists and where the gaps are. The needs analysis decides what interventions will close them. Diagnosis then prescription. Professional guidance explicitly advises keeping a clear distinction between the two to avoid duplicating work.

In practice the audit comes first and its output is the input to the analysis. Running them together tends to produce a document that recommends training before it has established a baseline, which is how organisations end up buying courses that address a problem nobody measured.

Why does the audit belong before the plan, not after?

Because the plan is supposed to be built on it, and reviewers can tell when it was not.

A good Workplace Skills Plan considers current and future needs, taking account of gaps identified through a skills audit, the performance management system, succession planning and any new process or technology planned for the year. That sequence is the intended one. The audit produces the evidence, the plan spends against it.

What happens instead, often, is that the plan is assembled from last year's plan plus whatever training people remember, and the audit is either skipped or produced afterwards to justify decisions already made. It still submits. It just does not tell you anything, and it is noticeably weaker when a discretionary grant application has to explain why this training and not some other training.

Start with a pilot

The most useful practical advice in the whole field is also the least followed. Before auditing the organisation, audit one unit with reasonably uniform roles.

You will find out how many reminders people need, whether the framework makes sense to the managers using it, whether the scale is being interpreted consistently, and how long it really takes per person. Then fix the instrument and go wide.

Design it to be repeatable while you build it. A one-off audit is a snapshot that starts ageing immediately. An audit you can run every year, from the same framework, with the same fields, becomes the thing your submission, your scorecard and your workforce planning all draw from.