The Transformation Triangle, the AI in L&D report from Donald H Taylor and Eglė Vinauskaitė, makes the case that content is no longer what makes L&D valuable (we offer our thoughts on that in this blog). It then offers three things that are. An L&D function can organise itself around tracking and closing skill gaps, around moving expertise between the people who have it and the people who need it, or around fixing the conditions that shape how well people perform. Each is a different claim on where L&D earns its place.
The report is careful not to rank them and makes one warning especially well, that we’d like to second. Teams tend to reach for the skill-gap route first, not because they’ve checked that the business needs it, but because building and tracking capability is what L&D has always done. Choosing by force of habit rather than by strategic need is how the wrong route gets picked.
Why teams default to the skill-gap route
The skill-gap route (the report calls it Skills Authority) is the natural default. Tracking capability and building programmes to close gaps is what L&D already thinks of as its job, so it feels safe.
The report’s caution is specific. That route only works where knowing exactly who can do what is operationally necessary, not just useful to have. Think of a pharmaceutical company that needs certified compliance, a consultancy that has to put the right expertise in front of the right client, or a global business staffing cross-border teams where personal networks aren’t enough. Where those stakes are real, skills intelligence is essential and L&D is the right home for it. Without them, you get the familiar dead end: the skills framework gets built, the platform gets bought, and employees ignore both because their pay, progression and projects don’t depend on what either one says.
So the real question isn’t “which role feels most like the L&D we already are?” It’s “which role does the business in front of us actually need?”
Start with the problem, not the end point
Pick the performance problem you’d want L&D to own if the function were set up differently. Not a learning problem. A business one. Weak client retention, maybe, or slow response to a shifting market, or no reliable read on who can staff a project, or risk building up because nobody sees a capability gap until it lands in the results.
Whichever problem you land on points you toward one of the three routes:
- The skill-gap route, if you can’t see what your people are capable of and it’s costing you. You track capability and close the gaps.
- The expertise route, if your strongest people know things that never reach anyone else. You surface what they know and move it to the people who need it.
- The conditions route, if people are good at their jobs but the way work is set up keeps getting in their way. You fix the workflow, tools and incentives around them rather than retraining the people.
Starting from a real, felt problem does something else useful. It gives you the business backing to push through the resistance that any of these shifts will meet.
Be honest about what your organisation will back
A matching problem isn’t enough on its own. What you’ll be allowed to do matters just as much. Some organisations will let L&D take on performance accountability and a hand in how work is designed. Others keep L&D firmly in the training lane and resist anything wider.
Reading that honestly tells you which boundaries you can push now, and which you’ll have to shift before you start. The conditions route (the report calls it Adaptation Engine) depends on this more than the others. It only works if the organisation will let L&D say uncomfortable things: that the problem is the incentive structure, or the tool, or a decision already made, rather than the people. Where that kind of honesty gets shut down, the route stalls no matter how sharp the diagnosis.
How to tell which route matches your situation
Track and close skill gaps. This is the route when putting the wrong person in the wrong role carries real consequences, and when what people can demonstrate already affects their pay, progression and the work they get. That last part is the catch. If none of those things currently depend on the skills picture, employees have no reason to engage with it, so you’d have to make development matter and build the tracking system at once. That is far harder than building the system alone.
Spread the expertise you already have. This is the route when the knowledge that makes the difference already sits in your people (your senior practitioners, your frontline, your client-facing teams) and builds up faster than any central team could ever write it down. The job here isn’t producing content. It’s surfacing what people know and moving it to where it’s needed. This is the route closest to our own work: connecting AI to an organisation’s own material, with each source traceable, so expert knowledge turns into reusable content. SkillFLO was built around that idea. Two things make this route work. People have to get something out of sharing what they know, whether that’s reputation, less repeated questions, or simply keeping up in a fast-moving field, because if sharing is pure overhead they won’t do it. And the workforce has to be spread out or varied enough that no single central curriculum could serve everyone well, which is what makes local expertise worth surfacing in the first place.
Fix the conditions around performance. This route needs two things. First, L&D has to be in the room early, before a problem has been turned into a training request, which usually comes from its reporting line or strong relationships with the business. Second, the organisation has to actually want the real cause found, even when that cause is an incentive, a tool or a past decision rather than a skills gap. If L&D only gets called in once someone has already decided the answer is training, neither condition is there, and you have to build that standing before this route is open to you.
You don’t need certainty to start
You likely won’t have a clear answer on day one, and you don’t need one. Begin where the report suggests: build relationships across the business and get close to the problems people actually care about. That work pays off whichever route you end up taking, so there’s no wrong place to start.
Worth saying plainly: some organisations aren’t ready for any of this yet. If leadership still sees L&D as training and compliance and isn’t open to more, these three routes are a future you can’t reach tomorrow. It would change the first question. Instead of which route to take, you start with what would have to shift for any of them to open up, and that groundwork is something L&D can begin building now.
