How to Design a Training Workshop That Changes What People Do

What does good workshop design actually mean?

A well-designed workshop changes what people do at work. That sounds obvious, but most workshops are judged on how the day felt: the facilitator's energy, the stories, the evaluation score. Those things matter, yet a pleasant day with no change back at work is an expensive day.

Time is also scarce. ATD's 2025 State of the Industry report, based on data from 539 organisations, found that formal learning hours per employee fell to 13.7 in 2024, from 17.4 the year before. When people get a dozen or so formal hours a year, each minute of a workshop has to be designed, not filled.

Good design is a sequence of decisions, each one made on purpose. The eight steps below follow that sequence, from the first conversation with the person who asked for the workshop to the moment someone else can run it without you.

What are the steps to design a training workshop?

The steps are not rigid. You will loop back when you learn something, especially after a pilot. But the order matters, because each step depends on the one before it. Objectives written before you know the business goal drift towards content. A session plan built before triage runs out of time.

How do you start with the business goal?

A request usually arrives as a topic: customer service, feedback, time management. Ask the sponsor for 45 minutes and turn the topic into a problem. What is happening now that should not be, or not happening that should? Can you see it in a number? What would people be doing differently in three months if this worked?

Write the goal as one sentence: a measure, from what, to what, by when. If there is no number, agree an observable behaviour and a date to review it. Then list who must do what differently. That list becomes the spine of your objectives.

Cathy Moore's Action Mapping, a well-known approach in corporate training design, works in this order: a measurable business goal, then the actions people need to take on the job, then practice activities for those actions, and only then the minimum information needed. Before any of that, it asks whether training is the right solution.

How do you know whether training is the answer?

Many performance problems are not skills problems. Before you design anything, check the other causes with the sponsor and with people who do the work. Clarity: do people know what good looks like and that it is expected? Environment: does a process, system or tool make the right action hard? Incentives: is anyone rewarded for the wrong behaviour? Memory: is the task so rare that a job aid would serve better? Skill: have people practised a hard skill with feedback? Motivation: do they know how but choose not to? Training mainly fixes the skill cause. If the main cause is a process, a tool or an incentive, say so in writing and name an owner for that fix.

Once you know a workshop is part of the answer, talk to three to five people from the audience and at least one of their managers, for about 15 minutes each. Ask about a normal day, what they find hard, what past training did for them and what would make a session worth their time. Write down misconceptions in their own words, note constraints such as shifts, time zones and languages, and ask about access needs. Each conversation should change at least one design decision.

How do you write learning objectives for a workshop?

Turn each on-the-job action into an objective that describes what people will be able to do by the end of the session. Use an observable verb, a condition and a standard: 'Choose the resolution option that fits the case and your delegated authority, using the options grid, for six scenarios, with a reason linked to the grid.'

Bloom's taxonomy, revised in 2001 by Lorin Anderson, David Krathwohl and colleagues, names six levels of thinking: remember, understand, apply, analyse, evaluate and create. Use it to check ambition. If the job needs judgement, an objective at the remember level will not get you there. Avoid verbs you cannot observe, such as understand, know, appreciate and be aware of. Replace them with explain, classify, diagnose, choose, write or demonstrate.

Keep to three to six objectives for a day. More than that and some will only be covered, not practised. Every objective needs practice with feedback, and evidence you could see in the room and later at work.

How much content should a workshop include?

Less than the expert wants. Experts see the whole map and find it hard to leave anything out. Cognitive load theory, developed by John Sweller from the late 1980s, explains why that hurts: working memory can only handle a few new elements at once, so extra information crowds out the thinking people need to build a skill.

Triage every content item. Is it needed to do the job? How often do people meet it, how costly is an error, and how difficult is it? Practise the must-do items in the room. Move should-know items to pre-work or a job aid, such as a checklist or decision table used at the desk. Cut or link the nice-to-know. Design the job aid before the slides, so the workshop teaches people to use the tool they will actually have.

How should you structure and time a workshop?

Most effective workshops follow a shape: an opening that connects people to the real problem, short content blocks each followed quickly by practice, application to people's own work, review points, and a close that turns learning into commitments. Breaks are part of the design.

Build in retrieval. In a 2006 study in Psychological Science, Henry Roediger and Jeffrey Karpicke found that students who practised recalling material remembered more two days and a week later than students who spent the time restudying, even though the restudiers felt more confident. Short quizzes, brain dumps and recaps are not filler; they are how memory is built.

Plan the clock minute by minute. Set your own rules for the longest input block, the maximum share of lecture, the longest stretch without a break and the share of time for practice, then check every block against them. Put practice or movement after lunch, never your longest input. If you run late on the day, cut input and discussion and protect practice and the close.

How do you design a virtual or hybrid workshop?

Redesign it rather than moving the same day onto a video call. In a 2021 analysis, Stanford researcher Jeremy Bailenson named four reasons video calls are tiring: intense close-up eye contact, constantly seeing yourself, reduced movement and the extra effort of sending and reading non-verbal cues.

Split a day into sessions of two hours or less with real work between them. Add a producer to run polls, breakouts and captions. Give people something to do every few minutes, and give breakouts of three or four a written task, a time and a shared output. For hybrid sessions, give remote people first turn in discussions and one shared digital board for everyone.

Build accessibility in from the start: ask about adjustments in every invitation, caption every clip, use readable materials and offer alternatives to physical activities. Many organisations use the W3C's WCAG 2.2 guidelines as the reference for digital materials. Rules on accessibility differ by country, so check your HR team and local law.

How do you make sure people use what they learn?

Design what happens before and after the workshop, not just the day. Support from managers is often weak: in the CIPD's Learning at Work survey 2023, only 51% of the L&D practitioners surveyed believed people managers encouraged participation in learning, and 39% said employees were given time away from their day-to-day role for it.

Brief managers two weeks before with three asks: a short chat before the workshop, a 15-minute check-in within a week, and watching the new skill in use around week six. Ask participants to write if-then plans at the close. Schedule spaced practice, such as short quizzes at two, four and eight weeks. Then pilot the whole design with an observer, fix the high-severity issues before launch, version your files and write a facilitator guide good enough for someone else to run it.

Decide how you will measure success before launch, using the measures agreed in the brief. Donald Kirkpatrick's four levels (reaction, learning, behaviour, results) are the best known. Will Thalheimer's Learning-Transfer Evaluation Model adds the useful warning that attendance and satisfaction say little about whether people can do anything new, and Robert Brinkerhoff's Success Case Method looks closely at the most and least successful participants to learn what helped and what got in the way.

If you want a structured place to do all of this, The Workshop Design Studio from Bodhih Training gives you a file for each step, including a session planner that checks timing, breaks and practice share automatically.