18 May 2026
Every skills-based L&D conversation eventually arrives at the same uncomfortable moment. The taxonomy is in. The managers are on board. The dashboards are saying something useful. And then someone in the room asks the simplest question of all: OK — so when an employee needs to actually learn one of these skills, what do we point them at? Nine times out of ten, the answer is a course catalogue built three to five years ago, organised by topic rather than skill, and last reviewed when half the current workforce wasn’t even hired.
That mismatch is the quiet bottleneck behind a lot of skills strategies right now. The strategy layer has moved on; the content layer is still 2018. Until that gap closes, every other investment — taxonomy, analytics, manager enablement, internal mobility — runs into the same wall: the learning the data points to isn’t actually there in a usable form. The next move in skills-based L&D, for most organisations, isn’t another platform decision. It’s a content decision.
Why the Old Catalogue Doesn’t Fit a Skills-Based World
Most L&D libraries were built around a clean assumption: people need to learn topics. “Time management.” “Effective communication.” “Excel for beginners.” Each course was a self-contained module, between thirty minutes and two hours, dropped onto a tile-based homepage and recommended by job role. It made sense when L&D was measured by completion rates and people genuinely did sit down once a quarter to take a course end-to-end.
A skills-based world breaks that model in three places at once. First, the unit of value is no longer the course — it’s the skill, at a specific level, in a specific context. Second, learners don’t arrive with two free hours; they arrive with seven minutes between meetings and a problem they need to solve. Third, the data that should be flowing back into the skills profile — what the person actually engaged with, what they tried in their work, where they got stuck — is invisible to a system that only knows whether someone clicked “mark as complete”. The catalogue isn’t wrong; it’s just answering an older question.
From Topics to Skills: A Different Unit of Content
The first shift is conceptual, and it’s the one most catalogues quietly fail. In a skills-based library, the unit of organisation isn’t a course — it’s a skill, tagged at a level. A single ninety-minute course on “Effective Feedback” becomes, in the new model, a small constellation: a five-minute explainer on the foundational skill, a scenario-based practice at the intermediate level, a manager-only debrief at the advanced level, plus the short reference content people actually open before a difficult conversation. Same content, fundamentally re-shaped.
That re-shaping is what unlocks everything downstream. When content is tagged to a skill and a level, the system can finally answer the question learners are actually asking — “I need to get from intermediate to advanced on this specific skill, in the next six weeks; what should I be doing?” The course catalogue couldn’t answer that. A skills library can. And the same tagging is what lets analytics start showing capability movement instead of completion volume.
Shorter, Closer to the Work, and Far Less Linear
The second shift is structural. Content for a skills-based world isn’t just re-labelled — it’s re-formed. The long, linear course is being unbundled into shorter, modular pieces designed to drop into the flow of work, alongside richer formats like scenario practice, peer-shared playbooks, short videos from internal experts, and step-by-step references that live next to the task they support. The point isn’t that long-form learning is dead; it’s that it can no longer carry the whole load on its own.
For L&D teams, this is where the work changes character. Rather than commissioning one new course a quarter and watching its completion rate drift downward over twelve months, the team is curating a much wider mix — some authored in-house, some pulled from an existing library, some captured live from how the best people in the business already work. KnowHow’s AI-powered course creation tools and course management capabilities sit naturally here — they make it realistic to author, tag, and maintain that kind of granular content without a content team five times the size.
Tagged to the Taxonomy, or It Doesn’t Count
The third shift is the least glamorous and probably the most important. Every piece of content has to be tagged to the skills taxonomy — the same taxonomy the rest of the skills programme is built on. If a video, a course, a job aid, or a scenario practice can’t be traced back to specific skills at specific levels, it can’t be recommended intelligently, can’t be measured against capability movement, and can’t be connected to the development goals managers and employees are agreeing on.
This is the moment a lot of L&D teams hit a wall — not because the work is intellectually hard, but because retro-fitting tags onto a legacy library is a slog. The way through is to stop treating it as a one-off cleanup project and start treating it as the standard way new content gets in. From now on, nothing enters the library without a skill, a level, a context, and an owner. Over a few quarters, the catalogue turns into a library — and the library starts to behave like an asset rather than an archive. Linking those development goals back to the work is exactly where features like training goals and insightful analytics & reporting become useful, because the tags finally have somewhere to land.
AI Doesn’t Replace the Library — It Demands a Better One
It’s tempting to assume that AI quietly solves the content problem — that a clever model will generate whatever a learner needs in the moment and the library question goes away. Generative tooling really has changed what’s possible. It can draft outlines, summarise long content into short reference cards, translate, role-play, and personalise tone. What it can’t do is invent the organisation’s point of view on how work actually gets done here. That has to come from inside the business, captured deliberately, and curated as content.
In practice, AI raises the bar on the library rather than removing the need for one. The better the source material — well-tagged, current, written in the organisation’s own voice — the better the AI’s recommendations, summaries, and adaptive paths get. The L&D teams getting genuine value from AI aren’t the ones with the flashiest model; they’re the ones whose underlying content is in good enough shape for a model to work with.
The Quiet Rebuild Behind the Strategy
Most of the leaders we speak to don’t describe the content shift as a project — they describe it as a slow, deliberate rebuild that runs underneath the bigger strategy. New content is born skill-tagged. Legacy content gets re-tagged when it’s touched, and quietly retired when it isn’t. Formats diversify; the catalogue stops being a wall of tiles and starts being a layered library that recommends differently depending on the skill, the level, and the person.
When that rebuild lands, the rest of the skills programme stops dragging. Managers can recommend specific things, not generic courses. Career and mobility conversations point at content that exists and fits. Analytics finally measures capability movement and not just clicks. The content layer becomes what it was always meant to be — not the archive at the end of L&D, but the working surface where strategy and people actually meet.
