9. Easier said than done
So far, we’ve discussed several reasons why shifting from learning to performance is difficult for L&D:
Believing that capability building is a more important driver of performance than designing work.
An over identification with learning versus performance.
L&D's structural location within the business.
So, we’re certainly up against it!
And unfortunately, we’re not out of the woods yet (to be fair, I never said this would be easy!).
Because I want to highlight something else about ‘designing work’ - it’s bloody difficult!
And not because concepts are complicated - in fact, the concepts are straightforward when you break them down (and we shall be doing so in later emails).
It's difficult because we’re operating in complex environments, with competing priorities, fragmented ownership, disconnected tech, undocumented workflows, years of accumulated workarounds, and critical knowledge trapped in people’s heads.
Let me give you an example…
I was recently chatting with a friend who wanted his team to use AI to reduce time spent on routine tasks. And so, they built an AI prompt library, hosted on an internal company site.
Not a terrible idea, until you observe what happened next…
They announced the library, delivered some demos, and expected people to navigate to the website to find the relevant prompt at the moment of need.
Spot the problem?
The prompts are not embedded in the flow of work.
And so, their use comes down to heroic effort. I have to remember to stop what I’m doing, find the prompt, and then use it correctly.
So, unless the prompts genuinely make my work easier, and unless I’ve experienced those benefits first hand, I won’t be motivated to use them.
Not to mention, there’s no guidance on how to use them. There’s no consequence if I don’t use them. There’s no feedback on whether I used them correctly. There are no examples of what the output should look like once I’ve used them. And there’s no guidance on how to evaluate the output quality.
Thus leading to inconsistent usage - some do, some don’t.
And one might say “well, if some people use them, it was worth the effort, right?”.
Maybe.
But we’re heading towards a future where performance depends on people and AI working together. In that world, organisations can’t afford for best practice to be optional. If a workflow, prompt, agent, or automation delivers better outcomes, the challenge won’t be building it. The challenge will be making sure it gets used.
Unfortunately, most organisations are not set up to help us embed best practice. And so, when our stakeholders ask us for help, we can only do so much.
What we can do:
add guidance on how to use the prompts
make it clear to what standard the output should be
provide best practise examples
add criteria to evaluate the AI outputs against
provide links to additional support
create reminders to use them
But can we integrate the prompts into the flow of work?
Depends.
Best case scenario would be the prompt can be embedded into the tool being used.
What’s more realistic is a clunky workaround.
At worst, we can’t do anything, and rely on regular reminders from leaders and heroic effort from the user.
And if the page is down, or if I’m under pressure to do the task quickly, or if I miss the memo about the prompt library, we have to accept AI won’t be used to a consistent standard.
So, what we’re advocating for when we talk about ‘redesigning work’ often requires changes we can’t easily make.
Damn!
So, what can we do?
First, there are ways to help. When we get really close to the work, and understand the way people operate, we can often figure out clever ways to build in best practices. They may not be optimal, but they can be enough to improve performance.
Second, as AI changes how we work, organisations will be forced to understand the implications of using rigid, legacy tools that don’t have the flexibility to integrate what people need at the moment it’s needed.
And they’ll need to take the idea of a performance guidance system seriously (a concept central to upcoming emails).
So, what we can do is make sure we’re prepared for these conversations - because they’re coming.
And by reading, reflecting on, and discussing the themes in these emails, you already are doing the work needed to prepare.
So, well done :)
Yours,
- Ant
