Training Needs Analysis: Best Practice for Finding the Real Gap
How to run a TNA that finds the real gap and gives leaders evidence — not a survey that just confirms what people already assumed.
Why most TNAs fail before they start
Most Training Needs Analyses fail for a simple reason: they start from an assumption that training is needed, and work backwards to justify it, rather than starting from the performance gap and testing what's actually causing it.
A TNA that begins with "what course do you think you need" has already skipped the one question that matters: is this a training need at all?
The questions a good TNA actually asks
A defensible TNA tests whether the knowledge or skill is genuinely missing, or whether performance is being held back by something else — unclear roles, weak governance, a structure working against the outcome, or simply unclear expectations.
It also asks what evidence exists for the current position, rather than relying on the loudest stakeholder's opinion of where the gap sits.
Evidence vs assumption
The difference between a TNA that leaders can act on and one that gets quietly shelved is almost always the evidence base. A TNA built on a handful of interviews with people who requested the training in the first place will tend to recommend more training — that's a sampling bias, not a finding.
Performance data, error rates, incident reports, and structured observation are all more defensible starting points than a survey of what people say they want.
Common mistakes
The most common mistake is letting a TNA become a wishlist exercise — asking stakeholders what training they'd like, then packaging the answers as a needs analysis. The second is failing to establish a baseline, so nobody can say afterwards whether the intervention actually worked.
A close third is treating every request as equally urgent, rather than prioritising by evidenced impact on performance.
What a defensible TNA looks like
The NATO and Royal Navy Training Modernisation case study is a clear example: a DSAT-compliant TNA pinpointed the specific points in the training pipeline where learners were being set up to fail, rather than recommending a wholesale redesign.
That precision — targeting the few points that actually move performance, rather than redesigning everything — lifted pass rates by 17% and cut failures by 20%, with far less wasted effort than a blanket response would have taken.
Common questions on this topic.
It depends on scope, but a focused TNA against a specific performance problem is typically weeks, not months. Scope creep — trying to analyse everything at once — is usually what turns a TNA into a multi-month project.
A mix of people closest to the performance problem and people accountable for the outcome — not only the people who originally requested training, whose view is useful but not sufficient on its own.
That's a legitimate and valuable finding — it means investment can be redirected to whatever is actually causing the gap, rather than being spent on training that wouldn't have worked.
Want this thinking applied to your organisation?
Insight is useful. Applied insight changes outcomes. Let's talk about yours.