Training Needs Analysis: How to Conduct One in 5 Steps

Last updated July 2026.

A training needs analysis is a structured way to find out where performance falls short of what the business needs, and whether training is actually the fix. It runs in five steps: organizational analysis, job or task analysis, person analysis, cause analysis, and a solution recommendation. Done properly it takes two to four weeks for a department and it usually ends with fewer courses than people expected, because a large share of performance gaps turn out to be process, tooling or motivation problems that no course can solve.

Most L&D budgets get spent on training that was requested rather than diagnosed. A manager says the team needs communication training, a course gets bought, everyone completes it, and the original problem is still there six months later. A training needs analysis (TNA) is the discipline that sits between the request and the purchase. Here is how to run one that gives you an answer you can defend to a CFO.

What is a training needs analysis?

A training needs analysis is the process of comparing the knowledge, skills and behaviors your organization currently has against the ones it needs, then deciding which of those gaps training can realistically close. The output is not a course list. It is a ranked set of gaps, each with evidence, a root cause, an owner and a recommended intervention, which may or may not be a course.

The distinction that makes a TNA worth doing is the difference between a can't-do problem and a won't-do problem. If a warehouse team is failing safety checks because nobody showed them the new procedure, that is a can't-do gap and training fixes it. If they know the procedure and skip it because the shift targets make following it impossible, training will change nothing and will annoy everyone. Separating those two is the single highest-value thing a TNA does.

The five steps of a training needs analysis

Step 1: Organizational analysis

Start at the top, with what the business is trying to do in the next 12 to 18 months. Entering a new market, launching a product line, absorbing a regulatory change or integrating an acquisition each imply capabilities the organization may not have yet. Read the strategic plan, the board deck and the last two quarters of operational reporting, then talk to the executives who own those goals.

You are looking for three things: the goals that depend on people doing something new, the resources available to support training, and the cultural conditions that will make training stick or fail. A company that pulls people out of training for urgent work has an organizational constraint, and no amount of content design gets around it.

Step 2: Job or task analysis

Next, map what each role in scope actually has to do and the knowledge, skills and abilities required to do it well. Job descriptions are a starting point and usually a poor one, since they tend to be written for hiring rather than for describing the work. Better sources:

• Interview two or three high performers in the role and ask what they do that others do not
• Shadow someone for half a day and write down the tasks nobody documented
• Ask supervisors which two or three capabilities separate a strong performer from an average one
• Pull the standard operating procedures, checklists and systems the role touches daily

The output is a task list per role with the underlying skills attached. This is the reference point everything else measures against, and it is also the thing that makes your eventual training assignable by role rather than sprayed across the whole company.

Step 3: Person analysis

Now find out who has the gap and how wide it is. Individual analysis is where most TNAs get sloppy, because self-assessment surveys are cheap and unreliable. People who are weakest at a skill are frequently the most confident about it. Triangulate instead:

• Assessment data: short scenario-based knowledge checks tied to the tasks from step 2
• Performance data: error rates, rework, escalations, quality scores, sales stage conversion
• Manager rating: supervisors rating each direct report against the same task list
• Self-assessment: useful for confidence and motivation, weak for actual capability

Where the manager rating and the assessment disagree sharply, that is worth a conversation before you build anything. It usually means either the task list is wrong or the manager is measuring something else.

Step 4: Cause analysis

This is the step that gets skipped and the step that saves the most money. For each confirmed gap, ask why it exists. The common causes fall into a short list: the person never learned it, the person learned it and forgot, the process makes doing it correctly slow or impossible, the tools do not support it, the incentives reward something else, or the wrong person is in the role.

Only the first two are training problems. If the gap is that the role is scoped wrong or under-leveled and every capable hire leaves within a year, a curriculum will not save it. Checking what the role pays in your market against what you are offering answers that question in an afternoon, and it is a cheaper answer than a training program that fails.

Step 5: Solution recommendation

Finish with a short document that a business leader can act on. For each gap: what it is, who has it, what it costs the business, what caused it, what you recommend, what it will cost, and how you will know it worked. Rank the list by business impact rather than by how easy each one is to build.

Where the recommendation is training, be specific about the format. A 20-minute assigned module with a knowledge check is a different commitment than a two-day facilitated workshop, and the TNA is where you justify the difference.

The three levels of training needs analysis

A training needs analysis template you can copy

You do not need software to run a TNA, only a consistent structure. One row per gap, with these columns:

• Gap: the capability that is missing, stated as an observable behavior
• Role and headcount: which roles are affected and how many people
• Evidence: the data that proves it, not the anecdote that suggested it
• Business impact: what it costs in dollars, risk, time or quality
• Root cause: knowledge, skill, process, tooling, incentive or fit
• Recommendation: the intervention, with format and length
• Owner and date: who is accountable and when it lands
• Success measure: the number that should move, and by when

Fill the evidence column before the recommendation column. If you cannot write anything in evidence, you have found a request rather than a need, and it belongs in a parking lot until data supports it.

How often should you run one?

A full organization-wide TNA is a heavy exercise and most companies run one every 12 to 18 months, usually timed just before annual planning so the results can shape a budget. Between those, run lightweight departmental analyses whenever something changes materially: a new system goes live, a regulation shifts, a product launches, or a team's performance metrics move in the wrong direction for two consecutive quarters.

Onboarding is the exception worth a standing analysis. New hires generate the clearest possible signal about where your training is thin, because they hit every gap in the first 90 days. If you track where new people get stuck, you get a rolling needs analysis for free. That data feeds directly into a stronger onboarding LMS program instead of waiting for the annual cycle.

Turning the analysis into training that actually ships

The failure mode after a good TNA is a 40-page report that nobody builds from. Keep the gap-to-course path short. Take the top three to five gaps by business impact, build a short module for each rather than a comprehensive curriculum, assign it to exactly the roles the person analysis identified, and set a due date. A narrow course that ships in two weeks beats a complete one that ships in two quarters.

The assignment logic matters as much as the content. If your task analysis says forklift operators need refreshed load-handling knowledge and shift supervisors need something different, those should be two assignments, not one course everyone sits through. A skills tracking system keeps the map from step 2 alive after the analysis ends, so the next TNA starts from a current picture instead of a blank page.

Then measure. Set the success metric from your template before launch, and check it at 30 and 90 days. If error rates, escalations or ramp time do not move, the gap was probably not a training gap, and that is a useful finding rather than a failure. Running all of this from one employee training platform means the assignment, the completion data and the follow-up assessment sit together, which is what makes the next analysis faster than this one.

Frequently asked questions

What is a training needs analysis?

A training needs analysis is a structured process for comparing the skills your organization has against the skills it needs, identifying who has each gap, and determining whether training can close it. It runs across three levels, organizational, job or task, and person, and ends with a ranked set of recommendations rather than a course list.

What are the 5 steps of a training needs analysis?

The five steps are organizational analysis (what the business strategy requires), job or task analysis (what each role must be able to do), person analysis (who has the gap and how wide it is), cause analysis (why the gap exists), and solution recommendation (what to do about it, with cost and success measures). Cause analysis is the step most often skipped and the one that prevents wasted spend.

What is the difference between a training needs analysis and a training needs assessment?

In practice the terms are used interchangeably. Where practitioners do draw a line, assessment refers to the broad identification of needs across an organization, while analysis refers to the deeper diagnostic work of confirming a specific gap and its root cause. Both describe the same goal: deciding what training is genuinely required before you build or buy anything.

How long does a training needs analysis take?

A focused departmental analysis typically takes two to four weeks: about a week gathering strategy and task data, one to two weeks on assessments and manager input, and a few days on cause analysis and the write-up. A full organization-wide TNA across many functions commonly takes six to ten weeks and is usually run once every 12 to 18 months ahead of budget planning.

Who should conduct a training needs analysis?

L&D or HR usually owns the process, but the analysis fails without line managers and a sponsoring executive. Managers supply the task detail and performance data, executives confirm which business goals depend on new capabilities, and L&D supplies the method and keeps the diagnosis honest when a stakeholder arrives with a solution already chosen.