The DUDUR Method

Why another method?

Most people who have worked with e-learning or learning development are probably familiar with the ADDIE model.

Addie stands for Analysis, Design, Development, Implementation and Evaluation.

ADDIE sounds good in theory, but in practice it always leads to a course, but not necessarily to a good result.

Why? Because we often forget to factor behavioral goals, motivation, and the target group’s real challenges into the process. Many end up producing e-learning “for the sake of e-learning,” but forget to ask the question:

Are we actually changing anything?

The premise of DUDUR is that learning is a means to an end, and the end is behavior.

If you can achieve the behavior without learning, do it.

Because learning is the hardest and longest path to behavior.

Compliance is also behavior. Compliance is people who don't reuse their passwords. People who report data breaches when they see them. People who do risk assessments.

Behaviour.

You can create learning materials that convey knowledge so that you appear compliant, or you can create learning materials and other solutions that specifically aim to change behavior.

This is where DUDUR comes into the picture – a method that combines different approaches (including behavioral design, Cathy Moore's action mapping and Nick Shackleton-Jones's 5Di).

What does DUDUR stand for?

DUDUR consists of five phases, named after the Danish words Definér, Undersøg, Design, Udvikl and Rul ud:

1. D – Define the behaviour
2. U – Investigate
3. D – Design
4. U – Develop
5. R – Roll out

It is a model that guides you from the very first, crucial consideration of the purpose of your learning effort (behavioral goals), through the practical development of the material and through to the actual rollout.

Below you will find an overview of each phase, including examples and typical pitfalls.

1) Define – The goal is behavior, not just knowledge

In the first phase, it's about setting the direction. Many of us are trained to think in terms of “learning goals,” but in DUDUR we prefer to work with behavioral goals. Why? Because learning often ends up as an “info-dump” if we only focus on what people need to know.

• Instead, ask: What should people do in practice?
• For example, if you have a GDPR course, you specifically define: “Employees must report data breaches within 24 hours.”
• If you work with occupational health and safety, it could be: “Employees must always wear safety glasses in situation X.”
• For an IT security course, you specifically define: “Employees should not reuse their passwords”

By defining a specific behavioral goal, it becomes much easier to keep track of what actually needs to be learned. At the same time, it provides a clear guideline: You know how you will measure whether the effort is working – namely, based on whether people perform the action you are aiming for.

Many people jump straight to the content (“You need to know everything about GDPR!”) and forget what the purpose is. So take the time to ask: What behavior do we want? Why do people need to do this? What are the consequences if they don’t?

DUDUR is based on the assumption that people want to do the right thing if it is easy and they can see a point in doing it.

DUDUR also relies on the assumption that "learning" is a word for work. Learning is strenuous. Learning is not necessarily something we want to do if we cannot see a benefit. The benefit is almost always behavior, or the possibility of behavior. For example, becoming more efficient, better, or having more time off.

If you can solve the problem without anyone having to learn anything, you should definitely choose it. Learning is the longest path to behavior.

2) Investigate – Know your target audience and their barriers

Now you know what you want people to do. But do they do it just because you tell them to? Hardly. Maybe they have a lot of practical, emotional, or knowledge-related barriers that prevent them from acting the way you want them to.

In the Investigate phase, you must uncover:

• Barriers: Why don't they wear the safety glasses?
• Are they uncomfortable? Do they not fit? Are they too far away?
• Emotional obstacles: Are people afraid of sanctions if they have caused a data breach themselves?
• Knowledge gaps: Do people not know what to actually report and how? Here, learning objectives are derived from action objectives.

Use interviews, observation, or AI tools like ChatGPT to find the most common misconceptions. But remember to talk to the right people: the target audience themselves, managers, colleagues. Try to elicit the unspoken truths (“What aren’t they saying out loud?”).

Sometimes you start with “knowledge”, or rather information, especially in compliance. But then the principle is that you “count backwards” so that you end up with an action goal. If you can’t find an action goal, the information doesn’t need to be included because it doesn’t solve a problem.

An example: “GDPR stands for General Data Protection Regulation and originates from the EU.”

Does the user need it in order to act? No. Therefore, it should not be included unless it is something the target group expresses a need for.

Example

Herning Municipality wanted managers to write better job postings. First they created a PDF guide, but no one read it. Then they held a course – few showed up. When they examined the managers’ daily routines, they found that managers typically only wrote job postings right after an employee quit at the end of the month. The solution? Have consultants ready to help that very week. A brilliant behavioral design move that emerged from examining reality.

What isn't in the room

Why did it work to have help ready in exactly the week the managers in Herning wrote their job postings? Because it was where they already were. But that raises a question anyone who commissions or designs a course should ask: how do you know where people are when you yourself sit somewhere else entirely?

Daniel Kahneman has a name for the limitation: what you see is all there is. The way I put it: we build a story out of whatever is close at hand, and we don't notice that something is missing. What isn't there doesn't get a low weight. It gets none at all.

It cuts both ways. Employees act on what is in the room when it matters: a screen, a colleague, a phone ringing, a shortcut that worked last time. The procedure on the intranet isn't in the room, so it isn't part of the choice either. And the person planning the initiative sees their own room: the policy, the course, the budget. The employee's room isn't in it.

That is why the first question in Investigate isn't “do they know?” but: what can they see where they have to act?

The lemonade stand

Nick Shackleton-Jones has described the same mechanism from another angle, in a LinkedIn post about pilots. An L&D team pilots something new, the feedback is good, they roll it out, and usage is disappointing. His explanation: people follow the path, and at work the path is the diary. Ask a pilot group for feedback by a certain date and you have put the thing in their path yourself. The pilot goes well. The rollout goes badly. His conclusion: design the path, not the lemonade stand.

For Investigate, that means not gathering people to assess something you have set up yourself. Find out where they already pass by: the screen they open anyway, the meeting they already have, the place where the task begins. That is where the solution needs to be.

Ask open questions

Ask an employee whether the course was too long and you get an answer about length. The question has already set the frame, and the answer can only move inside it. Ask only about length and you rebuild the length while the real cause stays untouched.

So the questions need to be open. Not “was it too long?” but “when is it easy, when is it hard, what would help you most?” Then the problem can show up where it is, not where you expected it.

Research has long shown that the frame shapes the answer. Schwarz et al. (1985) gave people two different response scales for the same question about how much TV they watched. With the low categories, people reported watching less, because the scale itself became a yardstick. Saris et al. (2010) found that answers to “do you agree that …” are of lower quality than answers to questions that ask about the matter directly.

In the conversation itself there are seven moves, and they all follow from the same frame.

The first comes before we have even sat down: no agenda. If you have decided the answer before you walk through the door, you will find it. It is what you see is all there is again, only in the person asking, and it is the most expensive mistake in the whole investigation and the easiest to make. The second also belongs before the conversation starts: no preparation. If someone asks whether they should prepare, I say no. Preparation produces the tidy answer, the one people think they ought to give, and we want the ordinary one.

When the conversation begins, I go first and tell them about something silly I do myself. When both people put something on the table, they like the conversation and each other more (Collins & Miller, 1994; Sprecher et al., 2013), and in my experience that is where the things people wouldn't otherwise say out loud start to come out.

Along the way, I ask them to walk me through how they work. Not “what's the problem?” but “show me what you do”: what do you do first, what do you weigh up along the way, why do you do it like that? And I watch before I ask. What people do and what they say they do are rarely quite the same, not because they lie, but because most of a working routine runs without anyone thinking about it.

At the end come the two moves that turn the conversation forward. The magic wand: “If you had a magic wand, what would you change?” It removes the filters of what is feasible, what it costs and who might be offended, and it comes from the miracle question in solution-focused therapy (de Shazer, 1988). And the scale: “On a scale from 1 to 10, how much of a hassle is it?”, and then “what would move it one step?” Two experiments have compared solution-focused questions with problem-focused ones and found more concrete action steps with the solution-focused ones, and in the more recent one the scale produced more than the magic wand (Neipp et al., 2016 and 2021). The magic wand opens up the wishes. The scale turns them into a next step.

The answers usually point to one of four places: safety, meaning, ability or effort.

3) Design – Choose the right solution

Once you know your goal (Define) and have identified the barriers (Investigate), you are ready to design the effort itself. And the point here is that it is not necessarily an e-learning course that you need to create. Sometimes a checklist, a poster or perhaps a practical behavioral adjustment (as in Herning Municipality) is the best solution.

• Use behavioral design: Remove friction and make it easy to do the right thing. Use Social Proof and other techniques from behavioral design.
• Use the “Turn it Around” format: Make the communication recipient-oriented. Start by engaging people in their own problems or challenges. Make the information you want to push something the recipients want to pull .
• Pre-mortem: Imagine the project has failed. What went wrong? Was there a lack of management support, were emails suspicious, is the target audience too busy?

Sometimes the barrier is actually that the target audience lacks knowledge. This is where you need to think about learning objectives. If the action goal is that they should be able to report a data breach when they discover it, we can derive two learning objectives:

1. What is a data breach?
2. How (and when) to report a data breach.

The misunderstanding about interactivity

Many people believe that e-learning will be better if there is enough interactivity (clicks, drag-and-drop, quizzes). But interactivity is only meaningful if it supports real reflection or allows for error. Otherwise, it might as well be passive video or text. Instead, focus on how to keep users' attention and make the message relevant to them.

4) Develop – Create the concrete material

Now you start producing. It could be an e-learning, an email series, a video, or something else entirely. Consider in this phase:

• Storyboard and script: What will you tell at the beginning to hook people? How will you use examples, cases or questions?
• AI tools: You can save a lot of time by using ChatGPT for text drafting.

You can divide Develop into two sub-phases: planning and execution. If you are doing e-learning, you start by writing a storyboard before developing it. If you are going to record a video with interviews, you do pre-interviews, write questions, etc.

Some people misunderstand “Design” and “Develop” and mix them up. But in DUDUR, the Design phase is where you decide what solution you want to create, while Develop is the practical part.

In designing the solutions, there is a focus on behavior, i.e. getting users to engage with the material. This means that it is largely recipient-oriented communication. In terms of communication, it is “borrowed” from copywriting, which traditionally aims to sell us something. Here, we are sold information and the recipients do not pay with money, but with their time.

5) Roll out – Implement and avoid pitfalls

The final phase is about getting the solution out to the target audience so they actually use it. It sounds trivial, but there are many things that can go wrong:

• Communication: How do you present this to the target audience? Are your emails worded so that they don't look like phishing attempts?
• Timing: When are people most motivated? Maybe right after they've been given a new task or a new colleague.
• Technical setup: Do links work, is the LMS login easy to use, and do users get the necessary support?

Here you can also think about behavioral design: Should a reminder appear at a certain time? Should you integrate a small “nudge” into everyday life?

But where is the evaluation?

You might be thinking, “What about evaluating the impact?” That’s a good point. DUDUR integrates evaluation along the way, especially in the first and last phases:

• In Define, you clarify how you want to measure whether the behavior changes.
• In Roll Out, you check whether the solution works in practice. More importantly, you observe whether the behavioral goal is being met.

Unfortunately, too many projects end with “a tick in the LMS,” but no real evidence that people are actually doing the new thing. DUDUR encourages you to keep going: Measure continuously, adjust if necessary, and find out if the behavior change is taking hold in reality.

That's why DUDUR works

The DUDUR method is not just a new version of the old learning models. Its greatest strength is the built-in behavioral design. Instead of considering e-learning as the goal in itself, we ask: “What exactly should people do differently?” Then we work in a structured way to remove barriers and make it easy for people to act. At the same time, we focus only on communicating the information that actually leads to action – which makes the content much more relevant to the recipients.

DUDUR consists of five steps:

1. Define the behaviour: Behavior, not learning goals.
2. Investigate the target group and reality.
3. Design the solution (which is not necessarily a course).
4. Develop the material (e.g. e-learning, videos, checklists).
5. Roll out and ensure real implementation.

If you are looking for a concrete way to create e-learning or training courses that really make a difference, the DUDUR method may be the common thread you are missing.

About 5Di – and why DUDUR is different

Nick Shackleton-Jones' 5Di model (Define, Design, Discover, Deploy, Drive) is well-known and has many good elements. I've taken a course with Nick in London myself and am a big fan. However, there are a few problems I often encounter in practice:

1. Length and budget
   5Di calls for a comprehensive investigation phase, where you spend time and resources finding “experiences” to enhance learning. In theory, it’s brilliant, but in practice, budgets and time can be too small – especially when it comes to compliance learning, which often runs on a tight operating budget. The result is that you either skip the investigation or try to create a “big experience” that no one has the money or time to implement.
2. Does not support compliance training
   The consequence of Nick’s thinking is that if you follow that slavishly, you end up in the situation that “you simply can’t do compliance.” In the real world, you have to do compliance training – there are legal requirements or internal policies that you can’t ignore. That’s why DUDUR is more friendly and realistic about compliance. It says: “We know you have to do something. So let’s shift the focus to the behavior and make it as effective as possible.”
3. Think, Feel, Do
   5Di talks about defining goals in terms of Think, Feel, Do. It sounds sensible to distinguish between what people think, feel, and do, but in the end, “Think” easily becomes “Know” – that is, knowledge – and can therefore end up with an information dump. In DUDUR we take an even more radical approach: We focus exclusively on what people should do. If there is an emotional or knowledge aspect, we consider it a barrier or driver for the behavior. In this way, we always start with “What action is the goal?” – and derive “Think” and “Feel” afterwards, if necessary.
4. Five D's are hard to remember
   5Di consists of five steps, all starting with D. It's a catchy name, but in practice, people easily forget what exactly each of the five Ds is about – and in what order.

With DUDUR I have tried to make it more colloquial and actually easier to remember, especially if you say it in Danish: “Du dur!” If you were to use it internationally, you can use the English variant GUIDE (Goal, Understand, Ideate, Develop, Execute).