The training catalogue for next year is signed off in the autumn. In March, a team lead asks for a topic that is not in it, because in the autumn nobody had heard of it. Until the next catalogue, she manages on her own: she asks around.
Behind that scene is a question many L&D teams are arguing about right now. Should the organisation decide what people learn, or should employees share and pull knowledge themselves? Omnora’s guide “Employee-driven Learning” works through the question in detail, with stumbling blocks and a roadmap for the transition, and you can download it free of charge. This article compares the two approaches and tries to give the older one a fair hearing.
What top-down planning gets right
The centrally planned catalogue has a poor reputation, and not all of it is earned. When an organisation has to make sure every employee has received the same instruction on health and safety, data protection or information security, planning from the top is simply the right form: content is checked once, everyone gets the same version, and at the end there is evidence an auditor will accept. The same holds for strategic topics leadership sets on purpose, such as a new business model or a change the teams cannot see coming yet. Nobody would expect a workforce to discover that kind of content on its own. AI is a case in point across Europe: Article 4 of the EU AI Act, which makes AI literacy an obligation, has applied since February 2025 (EU). Leaving that to chance is not a learning culture. It is a gap.
Then there is predictability. A catalogue can be budgeted, a trainer can be booked, and at year end you know what took place.
Where the catalogue misses everyday work
The weakness shows up somewhere else: in timing. A course happens when it is scheduled. The knowledge is needed when an error occurs, an unusual case lands on the desk, or a customer asks something the manual does not cover. Weeks often pass between the training date and that moment, and by then much of it has faded.
What employees do next is no secret to any manager. They ask a colleague. That works as long as the colleague is around. If she is on leave, in a meeting or at another site, the waiting starts. Panopto puts the cost at 5.3 hours per employee each week, spent waiting on answers that sit with another person. The catalogue has not failed here. It simply cannot answer a question it never knew about.
A side note. Almost every presentation on employee-driven learning shows the familiar model in which most learning happens on the job and through colleagues, and only a small part in formal training. We find the direction plausible. The exact percentages that circulate with it are closer to a rule of thumb than a measurement, though. So we leave them out.
There is a second break, too. The people who plan usually sit further from the work than the people who do it.
What employee-driven learning redistributes
Employee-driven learning reverses the direction. Knowledge moves from those who have it to those who need it, in short formats that are available at any time. L&D stays, but its role changes. Four things shift:
- Topics come from the teams, not from the annual plan.
- Knowledge arrives at the moment of need rather than on the training date.
- Experts capture their own knowledge, while L&D supplies templates and approves.
- Success is measured closer to the work, for example by which follow-up questions disappear. The article on measuring learning success shows what that looks like.
DPD shows that this works at scale. There, the business units build their own training and L&D steers instead of producing. 240 authors. 1,800 learning videos. 2.5 years.
The limits of learning from below
Convincing as it sounds, the approach has a weakness worth knowing before you start. Shared knowledge is only as good as the people sharing it, and not everyone shares equally. In almost every organisation, the first to volunteer are those who enjoy presenting, while the quiet experts with the most valuable knowledge hold back. Left alone, the collection ends up reflecting who is loud rather than what matters. Some topics nobody writes up voluntarily, because they are awkward, dull or sensitive. And content ages: a video about a process that changed a year ago does more harm than one that never existed.
So learning from below needs more stewardship than the classic catalogue. Just a different kind. Someone has to provide the infrastructure, organise approvals, spot the gaps and retire outdated material. Where that is missing, you get an academy with many entries and little trust; why academies often feel empty despite a good platform is covered in the article on building a corporate academy.
When to keep the catalogue
Not every topic belongs in the teams’ hands. Three cases clearly favour central planning:
- Mandatory training with evidence requirements, where wording and completeness are checked.
- Strategic topics whose need the teams cannot see yet.
- New fields where nobody in the organisation has the knowledge so far.
Even here, the catalogue does not solve everything. Training Magazine reports that 49 percent of staff skip mandatory courses or barely engage with them. Planned centrally does not mean learned.
Start the shift small
If you are coming from a catalogue, do not begin with a platform for everyone. Begin with a group that wants to try. The guide explicitly recommends a small community rather than individual ambassadors, because one person on their own struggles to convey the benefit, whereas five who show each other what works do not. That group experiments, gives feedback in regular rounds, and later becomes the people who answer their colleagues’ first questions. Meanwhile HR and L&D explain why things are changing. It sounds obvious and is still the step most often forgotten. How the first weeks of such a rollout usually unfold, including the sluggish phase, is described in the article on expert generated content.
The next step is smaller than it sounds. Take next year’s catalogue and mark every line that is a genuine obligation or a deliberate strategic choice. Whatever is left is a candidate for the teams.




