How to run a training needs analysis across organization, task, and individual levels — find real gaps, prioritize, and turn it into a plan.
Got an LMS decision on your plate?
45-minute call. Plain-English audit. Fixed-price quote if there's a fit, or a "no" if there isn't. No deck. No pitch.
The four L&D trends that matter in 2026 — and what each one means for an operationally complex, multi-site firm that wants to adopt them without per-seat penalties.
Becoming a skills-based organization means mapping work to skills instead of titles, and your training platform has to support the shift.
A practical guide to measuring training effectiveness past completion, using Kirkpatrick's four levels and owned data.
A training needs analysis is the process of finding the gap between how people perform today and how the business needs them to perform — then deciding which of those gaps training can actually close. Done well, it stops you from buying courses nobody needs and points your budget at the problems that move the numbers. This guide walks through how to run a training needs analysis at three levels, where to get the data, how to prioritize, and how to turn the findings into a plan.
It is the diagnostic step that should come before any platform or content decision. If you skip it, you end up reverse-engineering a need to justify a tool you already bought.
A training needs analysis (often shortened to TNA) answers three questions in order:
That second question matters. The classic discipline here is to separate a genuine skill gap from a performance problem with another root cause. Training a team on a process they already understand but cannot follow because the system is slow wastes money and erodes trust in L&D.
The most durable model, drawn from decades of instructional-design practice, looks at needs at three levels. Run them top-down so each level constrains the next.
Start with the business. What goals, risks, or changes are driving the need? A new plant opening, a regulatory deadline, a turnover spike, an ERP rollout, an expansion into a new state with different compliance rules. The output here is a short list of business-driven priorities — not a course list.
For mid-market, multi-site firms this level also surfaces variation. A Texas facility and an Ohio facility may face different OSHA-driven hazards and different state training mandates. The organizational level is where you decide which needs are company-wide and which are site-specific.
For each priority, break the job down. What does competent performance actually require — which tasks, to what standard, under what conditions? This is where you define the target. A "forklift operator" need becomes a specific set of behaviors and a certification cadence rather than a vague "more safety training."
Useful sources at the task level include job descriptions, standard operating procedures, observation of top performers, and regulatory standards such as the relevant OSHA requirements or FDA Food Code provisions for food operations.
Now compare actual people against the standard. Who can already do the task, who cannot, and who is close? The gap at this level is the real training need. The mistake to avoid is starting here — running a skills survey before you know what the business needs or what good looks like just produces a popularity contest of training requests.
A credible analysis triangulates several sources rather than relying on one survey. Mix quantitative signals with direct observation and conversation.
Existing system data is the cheapest place to start. If your platform already centralizes completion records, assessment scores, and certification expiry across sites, you can spot gaps without launching a single survey. That is one practical argument for owning a connected platform rather than stitching reports from several disconnected tools — see our build vs buy guide for how that tradeoff plays out.
For each priority, write the gap as a plain sentence: current state, required state, and the size of the difference. For example: "Maintenance technicians at three of five sites cannot complete a lockout/tagout procedure to standard, based on observation and a recent near-miss."
Then apply a single filter before you call it a training need. Ask: if these people were paid a large bonus to do the task correctly today, could they? If yes, the problem is not knowledge or skill — it is motivation, process, or environment, and training will not fix it. If no, training is at least part of the answer. This one question prevents most wasted training spend.
You will find more gaps than you can fund. Prioritize on two axes: business impact and feasibility.
A simple high/medium/low score on each axis is enough. Fund the high-impact, high-feasibility gaps first, schedule the high-impact but harder ones deliberately, and park the rest with a documented reason. The skills-based view of the workforce can sharpen this further — our piece on the skills-based organization covers how to frame gaps as capabilities rather than courses.
The analysis is only useful if it becomes a plan with owners and dates. For each prioritized gap, define the objective, the audience by role and site, the delivery method, the owner, the timeline, and how you will measure success. Decide measurement up front — our guide on measuring training effectiveness explains why retrofitting metrics after launch rarely works.
Use this table as the working document. One row per gap; fill it during the analysis and hand it to whoever builds the program.
Anything that does not fit a row on this template is not yet ready to fund.
For a single business priority across a few sites, a focused analysis can take two to four weeks — most of it spent gathering and triangulating data rather than writing it up. A company-wide annual analysis takes longer because of the organizational-level coordination. Resist the urge to skip levels to go faster; a shallow analysis costs more later in wasted programs.
L&D usually leads, but it cannot be done in isolation. The organizational level needs leadership input, the task level needs subject-matter experts and frontline managers, and the individual level needs honest manager assessments. Treat it as a cross-functional exercise with L&D as the facilitator and owner of the output.
A skills gap analysis is one part of a training needs analysis — specifically the individual level, comparing people to a standard. A full TNA also includes the organizational level (why it matters to the business) and the root-cause filter (whether training is even the right fix). Running only a skills gap analysis risks treating every gap as a training problem.
That is a successful analysis, not a failed one. If the root cause is process, tooling, or staffing, document it and route it to the right owner. L&D earns credibility by declining to train away problems it cannot fix — and by spending its budget only where training genuinely moves the outcome. For broader context on where L&D priorities are heading, see our 2026 L&D trends and reference points from ATD.