breed — digital learning experts
breed — digital learning experts

E-learning agency based in Munich. Award-winning since 2014.

How desirable difficulties work in e-learning

Why challenging learning tasks can help, and why a hard-to-use interface isn’t one of them.

Tobias BertenbreiterUpdated: November 19, 2026
Desirable Difficulties statt Binge Learning im E-Learning

What are desirable difficulties, and why do they matter so much in e-learning?

Desirable difficulties in e-learning are learning conditions that make learning harder in the short term but, when the circumstances are right, can improve it in the long run. Sounds counterintuitive? It is. Rereading (going over the same material again and again) and cramming feel easy and familiar. Yet research on long-term retention shows clear advantages for retrieval practice and spaced practice over simply rereading or cramming at the last minute. Studies show that difficulties, used wisely, can improve retention and support transfer to new tasks. That’s especially relevant for digital and blended learning environments, where adaptive learning platforms let you plan repetitions and task sequences deliberately.

How difficulty helps learning: the theory behind desirable difficulties in e-learning

Bjork & Bjork’s New Theory of Disuse distinguishes between two key dimensions of memory: retrieval strength (how easily something can be recalled) and storage strength (how deeply it is anchored). Here’s the catch: information that comes easily (high retrieval strength) is often overrated. It feels familiar, but that says little about whether it will still be available later. This is exactly where desirable difficulties come in: learning tasks should challenge learners and make retrieval harder to achieve deeper encoding. At the same time, the level of difficulty must not become demotivating, and there is no universal ideal error rate. The often-cited 15% comes from Wilson et al.’s modeling study on the “Eighty Five Percent Rule” under specific learning conditions. It is not a universal target for web-based training.

Using desirable difficulties strategically in e-learning

Digital learning environments offer unique ways to put desirable difficulties to work. Here are three examples. 1) Spacing: content is repeated at intervals, with the gaps based on prior knowledge and the learning goal. 2) Interleaving: instead of presenting content in blocks, you mix topics, so learners have to consciously work out similarities and differences. 3) The testing effect: regular low-stakes quizzes give learners practice at retrieval and strengthen it. These strategies can support long-term retention. To get a more realistic picture of their own progress, learners should compare what they expected with their actual results.

The underrated power of self-assessment in learning

One often overlooked benefit of desirable difficulties is metacognitive. The illusion of competence, overestimating what you know and can do, shows up precisely when learning methods feel easy. Learners mistake ease for mastery. Difficult but solvable tasks, on the other hand, lead to more realistic self-assessment. In e-learning, you can support this with visual analytics, prompting techniques, or performance dashboards. Learners get feedback on their performance and can check their self-assessment against it. Keep in mind that a dashboard only reflects the tasks it tracks, not everything someone can do on the job.

What happens when desirable difficulties are used the wrong way?

Not every difficulty helps learning, which is why the difference between desirable and undesirable difficulties is so important. Convoluted wording, irrelevant extra information, or excessive time pressure certainly make things harder, but they do nothing for retention or transfer. What counts is the functional link between the difficulty and the learning goal. In e-learning, that means tasks have to be challenging but solvable. When users are repeatedly frustrated or overwhelmed, motivation drops and learning fails. Adaptive elements can help: platforms that adjust the level of difficulty dynamically can keep overload in check and offer the right level of challenge.

Key strategies for building desirable difficulties into digital learning

  • Spaced repetition: An algorithm calculates when content should come up again (for example, SuperMemo).
  • Contrasting cases: Interleaving by deliberately mixing different concepts.
  • Generative tasks: Tasks that require learners to actively produce a solution (for example, fill-in-the-blank exercises or concept maps).
  • Retrieval prompts: Questions designed to trigger recall, even when nothing is graded.
  • Well-timed feedback: Resolve mistakes in a way learners can understand. Whether feedback should be immediate or delayed depends on the task and the learning goal.
  • Explicit metacognition: Users track their learning progress and reflect on it through performance visualizations.

Desirable difficulties as a design principle for the future of digital learning

Several of these strategies are well supported by evidence on long-term retention. How well they work, however, depends on the task, prior knowledge, and the support learners get. For instructional design, that means we need clear, accessible user interfaces and deliberate challenges within the learning task itself. What’s missing isn’t technology. It’s the conscious instructional design decision to create the right learning challenges. The e-learning of the future builds in controlled challenges that deepen learning: not through more content, but through better timing, well-calibrated difficulty, and smart feedback structures.

Strategic outlook: where the research opportunities lie

Despite compelling evidence, key research questions remain open. For example: How can desirable difficulties be scaled in fully asynchronous learning environments? What role does personalization play, and how can difficulty be adapted to individual learners? And how do different types of feedback affect storage strength over longer periods of time? This is where the interplay of learning analytics and adaptive design holds enormous potential. Educational technology is entering a new phase: moving from distributing content to designing learning processes that work, using targeted challenges.

Our approach to desirable difficulties

At breed.design, we apply Bjork’s findings to our digital learning design. Beyond the principles above, one example is that we rarely build content that is purely video-based. In our e-learning courses, we always try to combine video with interactive content and tasks. That makes it “hard” for learners to simply consume their way through the material (we call this “binge learning”: taking in large amounts of content without actively processing it). Learners have to actively work through the content themselves. That’s how we aim to support long-term retention.

Sources and further reading

Tell us what you need. We’ll find the right solution.

Whether it’s a single course, a learning journey or a learning platform like notch.learn, our first call is about understanding what you need and how we can help.