In a world of huge rising demand, reskilling pressure and AI acceleration, L&D keeps saying it must be strategic. But operational data shows execution is where things fail: prioritisation, flow, capacity and measurement. Ryan Austin argues the real shift is operating better, not doing more, and lays out the evidence.
Over the past decade, the role of learning and development has been steadily redefined. L&D is no longer expected to simply deliver training. The expectation now is far broader and far more demanding. Teams are being asked to:
- Align directly to business priorities
- Enable workforce transformation
- Support reskilling at scale
- And, increasingly, demonstrate measurable business impact
At the same time, the learning market continues to grow, with investment expected to accelerate through 2030 as organisations respond to rapid technological change and shifting skill demands. Reports from LinkedIn and the World Economic Forum consistently highlight the urgency of reskilling, with many organisations expecting a significant portion of their workforce to require new capabilities within the next five years.
One phrase has become almost universal across L&D
Layer AI into that, and expectations rise even further. Faster delivery. More personalisation. Better outcomes. And in response, one phrase has become almost universal across L&D:
“We need to be more strategic.”
But there’s a tension building beneath that statement. Because while the ambition is clear, the operating reality often tells a different story.
Strategy is easy to say but execution is harder
There’s nothing wrong with the strategy conversation itself. In fact, it’s long overdue. For years, L&D has been pushing to move beyond order-taking and into a more influential role within the business. And many teams have made real progress:
- Stronger alignment with stakeholders
- Better articulation of goals
- More involvement in upfront discussions
But strategy doesn’t live in alignment meetings or slide decks. It shows up in the system.
In how work is requested. In how it’s prioritised. In how it moves. In how it’s measured. And this is where things start to break down.
What operational data reveals
Recent research conducted by Cognota, based on aggregated learning operations data across anonymised organisations, highlights a consistent pattern.
On the surface, L&D teams appear highly active:
- Thousands of requests flowing into the function
- The majority of those requests being approved
- Large volumes of projects and tasks created to support delivery
At first glance, this looks like maturity. Like scale. Like a function embedded in the business. But the underlying signals suggest something else.
- Completion rates hover around 50%
- Projects are almost universally reported as “on track”
- Work is not consistently managed through active stages
Taken together, this points to a system that is absorbing demand effectively, but struggling to convert it into consistent, predictable output. And that’s where execution risk begins.
The execution gap: where strategy meets reality
The gap between strategy and execution in L&D isn’t usually intentional. It emerges gradually. Teams adopt the language of strategy:
- Business alignment
- Outcomes
- ROI
But the underlying operating model remains largely unchanged:
- High approval rates
- Limited prioritisation discipline
- Inconsistent tracking of effort and progress
- Measurement as an afterthought
Over time, the gap widens. And eventually, it creates a situation where L&D appears strategic from the outside but operates reactively on the inside. That’s not just a perception issue. It’s an operational risk.
Five execution risks that are easy to miss but hard to recover from
These risks don’t tend to show up in planning sessions. They emerge in the flow of work itself.
1. Prioritisation risk: when everything gets in, nothing stands out
One of the clearest signals in the data is the volume of approved work. When most incoming requests are accepted, it usually reflects a well-intentioned effort to support the business. But it also suggests that prioritisation is happening informally or not at all. Without clear criteria:
- Work is added faster than it can be delivered
- Strategic initiatives compete with tactical requests
- Teams are forced to make trade-offs during execution, not before
Research in portfolio management has long shown that limiting work-in-progress is critical to maintaining throughput and focus. When that discipline is absent, systems become overloaded. What starts as responsiveness quickly becomes congestion.
2. Signal risk: when reporting stops reflecting reality
The near-universal “green” status across projects is another telling pattern. In any complex system, variation is expected:
- Some work will be ahead
- Some will be at risk
- Some will be delayed
When everything appears on track, it’s worth questioning the signal itself. Inconsistent definitions, infrequent updates, or lack of accountability can all lead to reporting that looks stable but isn’t. The risk here isn’t just inaccurate reporting. It’s delayed awareness. By the time issues become visible, they’re often already impacting delivery.
3. Flow risk: when work starts faster than it finishes
High task volumes are often interpreted as productivity. But volume alone doesn’t tell the full story. When completion rates lag behind creation rates, it suggests a breakdown in flow:
- Work accumulates in early stages
- Bottlenecks form in review, approval, or coordination
- Teams struggle to maintain momentum across multiple initiatives
This aligns with broader findings in knowledge work and product development: excessive work-in-progress reduces efficiency and increases cycle time (Reinertsen). In practical terms, it means that teams feel busy, but delivery becomes less predictable.
4. Capacity risk: when effort is invisible
Capacity is one of the most discussed and least understood areas in L&D. Most leaders know their teams are stretched. Fewer can quantify:
- How much work is actually being done
- Where time is being spent
- What trade-offs are required to take on new initiatives
Without that visibility:
- Planning becomes reactive
- Commitments are made without a clear understanding of impact
- Scaling delivery becomes difficult
This is where strategy quietly breaks down. Because strategy depends on making choices. And you can’t make effective choices without understanding capacity.
5. Measurement risk: when impact becomes anecdotal
There’s broad agreement across the industry that L&D needs to demonstrate impact. CIPD, ATD, and others have consistently emphasised the importance of moving beyond completion metrics toward business outcomes.
But in practice, measurement often lags behind execution. Not because teams don’t care, but because they don’t have the capacity to prioritise it. When measurement is inconsistent:
- All work appears equally valuable
- Stakeholders rely on urgency rather than impact
- L&D loses its ability to influence decisions upstream
Without that influence, it becomes harder to operate strategically at all [TJ agrees, and we have a report on it! -Ed]
The compounding effect
These risks don’t operate in isolation. They reinforce each other.
- Weak measurement undermines prioritisation
- Poor prioritisation increases workload
- Increased workload reduces capacity for measurement
- Limited capacity leads to inconsistent execution
Over time, this creates a cycle that is difficult to break. And it explains why many L&D teams feel stuck:
- Busy, but not progressing
- In demand, but not in control
- Strategic in intent, but operationally constrained
Why AI won’t solve this (on its own)
AI is often positioned as the solution to many of these challenges. And it will help, particularly in areas like workflow management, summarisation, and prioritisation. But there’s a risk in assuming AI will fix underlying operational issues.
Because AI tends to:
- Expand what’s possible
- Increase speed
- Lower barriers to business outcomes
Which, in turn, can increase demand even further. If the operating model doesn’t change, AI doesn’t remove constraints. It amplifies them. Some early evidence even suggests that heavy AI users aren’t working less, they’re just taking on more (BBC).
What’s actually changing in leading teams
The organisations starting to address this gap aren’t necessarily more advanced in terms of strategy. They’re more disciplined in how they operate. A few shifts are becoming more common:
Being explicit about trade-offs: Not everything gets done. And that’s acknowledged upfront.
Defining execution standards: Clear definitions of status, progress, and success, so signals can be trusted.
Making capacity visible: Understanding effort at a role and team level, not just at a project level.
Starting small with measurement: Focusing on a few high-impact programs rather than trying to measure everything.
Using AI to reduce friction: Not just to produce more but to streamline coordination, reporting, and decision-making.
None of this is particularly new. But it’s often been overlooked in favour of more visible initiatives.
Closing the execution gap: creating capacity where it doesn’t exist
If the problem is execution, the natural question becomes: How do you actually create capacity in a system that already feels maxed out? Because for most teams, capacity doesn’t feel constrained, it feels impossible.
- Headcount isn’t increasing
- Demand isn’t slowing down
- Expectations are only rising
So, the answer can’t be “do less” in a vacuum. It has to be how work gets done differently. And this is where the model starts to change.
1. Treating capacity as a system problem, not a people problem
One of the biggest mindset shifts is this: Capacity isn’t just about how many people you have. It’s about how your system operates. The LearnOps model itself reflects this, spanning Align, Plan, Execute, Measure, and Optimise as an interconnected system rather than isolated activities. In practice, that means:
- Reducing unnecessary work-in-progress
- Standardising how workflows
- Eliminating friction between stages
The goal isn’t to push people harder. It’s to make the system easier to execute within.
2. Moving from individual productivity to “augmented execution”
For years, L&D has focused on building the skills of people. That’s still important. But something new is emerging: It’s no longer just about the skills of your team. It’s about the capabilities of the system they operate within, including AI. The most effective teams are starting to embed AI directly into their workflows, not as a tool on the side, but as part of how work gets done.
This includes:
- AI agents that understand internal playbooks, methodologies, and standards
- Agents that can draft, review, or structure work based on business context
- Systems that can summarise requests, suggest prioritisation, or flag risks early
In many ways, this is the first time L&D has had the ability to scale execution without scaling headcount. The difference is subtle but important. This isn’t about using AI to generate more content. It’s about using AI to reduce coordination, decision-making, and execution friction. In other words, freeing up human capacity without increasing headcount.
3. Building “flash teams” instead of fixed capacity models
Another shift is happening in how work is staffed. Traditionally, L&D teams are structured around fixed roles and static capacity. But as demand becomes more variable, that model starts to break. Instead, some organizations are experimenting with more dynamic approaches, often referred to as “flash teams.”
These are:
- Temporary, purpose-built teams assembled around a specific initiative
- Blending internal roles, external partners, and increasingly AI agents
- Designed to deliver outcomes quickly, then disband
This approach:
- Reduces long-term capacity constraints
- Allows specialisation to be applied where it’s needed most
- Prevents teams from being permanently overloaded with ongoing work
It’s a shift from owning capacity → orchestrating capacity.
4. Creating operational guardrails (not just guidelines)
Many L&D teams already have processes. What’s often missing is enforcement. The difference between teams that scale and those that struggle isn’t whether they have frameworks, it’s whether those frameworks actually shape behaviour.
That includes:
- Clear intake criteria that limit what enters the system
- Defined status definitions that make reporting reliable
- Agreed thresholds for workload and prioritisation
Without these guardrails, even the best strategy gets diluted during execution. With them, teams can:
- Make faster decisions
- Protect capacity
- Maintain focus on high-impact work
5. Designing for flow, not just output
Finally, there’s a shift in how success is defined. Many teams still focus on output:
- Number of courses
- Number of programs
- Volume of work completed
But leading teams are starting to look at flow instead:
- How long work takes to move through the system
- Where bottlenecks occur
- How consistently work gets delivered
Because ultimately, capacity isn’t just about how much you can produce. It’s about how reliably you can deliver.
The real shift: from doing more to operating better
None of these changes are easy and they don’t happen overnight. But they point to a broader shift that’s starting to take shape:
From:
- Adding more work
- Adding more tools
- Adding more pressure
To:
- Redesigning how work actually happens
Because for L&D, the constraint is no longer content. It’s execution. And the teams that recognise that and act on it won’t just keep up with demand, they’ll finally be in a position to shape it.
The risk isn’t lack of strategy, it’s misalignment
L&D doesn’t lack ambition. If anything, it’s one of the most forward-looking functions in the organization. The challenge is alignment between what L&D says it is and how it’s set up to operate. Because strategy isn’t defined by intention, it’s defined by execution; repeated, consistent, and visible over time.
Right now, the data suggests that many teams are still working toward that alignment. And until they get there, the biggest risk isn’t that L&D won’t become strategic. It’s that it will continue to sound strategic, while operating in a way that makes that strategy difficult to deliver.
Ryan Austin is Founder and CEO of Cognota.
