Training needs analysis

The three levels

A proper TNA looks at the same question from three heights, and skipping the first is the usual reason a training budget gets spent on things nobody needed:

How do you do one?

1. Define required capability first. Before looking at anybody, write down what the work demands. A competency framework does most of this for you; without one you will end up comparing people to each other instead of to a standard.
2. Assess what you have. Use evidence rather than opinion where you can - see the sources below.
3. Name the gaps, and be specific. "Better at spreadsheets" is not a gap. "Cannot build the monthly margin report without help, so it stops when one person is on leave" is.
4. Prioritise honestly. Legal and safety obligations first, then gaps blocking something the business has committed to, then development that retains people you would hate to lose, then everything else.
5. Choose the intervention last. A course is one option among several, and often the weakest.
6. Decide the measure before you spend. What will be different, observably, in three months?

Where does the evidence come from?

Most of it already exists - the analysis is mainly a matter of reading it together rather than collecting something new:

• 1:1 and review notes - the same difficulty recurring across several people is a training need, not a personal failing. Regular 1:1s make this visible months before a review would.
• Errors, rework and complaints - where the same mistake keeps happening, someone was never actually shown.
• Exit interviews - "no development" is one of the most common reasons given for leaving. The exit interview form asks it directly.
• Escalations to one person - a queue outside one desk is a capability gap wearing a disguise.
• Anything changing - new systems, new legislation, a new service. These are needs you can see coming, which makes them the cheapest to meet.
• Absence and handover pain - what stops when someone is off tells you exactly where the knowledge is thin.

What counts as training?

Booking a course is the default answer and often the least effective one, because the knowledge arrives detached from the work and decays before it is used. Cheaper and usually better:

• Doing the work with someone experienced - the most reliable transfer there is.
• Writing it down. Half of what looks like a training need is an undocumented process. The fix is a page, not a workshop.
• Deliberate rotation - the only real cure for single points of failure.
• A stretch task with support - capability built on real work, where it will actually be used.
• External training - right for regulated content, formal qualifications, and genuinely new specialisms.

How do you know it worked?

Attendance is not a measure, and neither is a satisfaction score at the end of a session - people rate enjoyable training highly whether or not it changed anything. Set the measure when you identify the gap, in the terms the gap was described in: three people can now produce the margin report unaided; rework on that process has stopped; the queue outside one desk has gone.

Does a small business need one?

It needs the thinking, not the paperwork. Under about twenty people this is a spreadsheet and an afternoon: roles down one side, the capabilities the work requires across the top, an honest mark in each cell, and a hard look at every column with only one confident name in it. That single view usually reveals more than a formal process does - because the question it answers, what would we struggle to do if this person left tomorrow?, is the one that actually keeps people awake.