Businesses are facing an increasingly overwhelmed workforce, escalating token costs and a lack of evidence that AI is delivering ROI. In part one of this series, I explain why enterprise AI can’t succeed without first giving employees a baseline in workflow capability.
AI has given the workforce a superpower - with AI, we now have the potential to do more, and to do it faster.
But this comes with a very large, and very inconvenient caveat:
People can only create value with AI if they understand how they do their work.
The flawed assumption most of us are making
Now, this raises a common objection: “of course people understand how they do their work, how else are they getting things done?”.
But this involves a dangerous assumption - that people have defined ways of working.
After having led and observed several enterprise AI rollouts, I have seen a recurring theme - businesses assume their people operate using best practices. And therefore, we can simply feed those best practices into AI and increase productivity.
And that would be true if people did operate using best practices.
But they don’t.
Work is messy. We deal with broken processes, undocumented work, disconnected systems, competing priorities… Humans have been compensating for ambiguity at work for decades.
Which means the vast majority of work is invisible. It’s undocumented and lives in the heads of those doing it (we know this to be true because most of an employees knowledge disappears when they leave the company).
And often, even the people doing the work don’t have access to that knowledge either! Because the work is so ingrained in their subconscious, articulating what they’re doing (to the level of detail needed for someone else or AI to replicate) is a skill they don’t have.
Doing the work ain’t the same as defining it
Don’t believe me?
Let me share an example. How confident are you with your phone? I’d imagine you could check the cinema listings, book Gold Class tickets and order an Uber there without too much fuss, right?
Now think back to a time when you’ve tried to teach someone else how to use their phone. Maybe an elderly relative or a less tech-savvy friend. Was articulating what to do as easy as doing it yourself?
If you’re anything like me, this is tricky.
But whether it’s teaching someone how to use the phone, or trying to articulate something else built into your muscle memory, I’m sure you’ve experienced this phenomenon.
And so, being able to do something is entirely different to being able to articulate how you’re doing it - especially to a level of specificity someone else (or AI) could repeat.
This is a skill - it’s a skill that very few people have. And it’s a skill that requires practice, guidance and support.
Why workflow capability comes before the tech
So, this is where workflow capability comes in. We can give all our employees AI and ask them to use it. But that will inevitably lead to frustration.
Some will figure it out. There will be pockets of excellence.
But most will use AI on the fly, writing half-baked prompts as and when they need them. Which might make them incrementally more productive, but ultimately requires everyone to figure things out as they go, and will lead to different results for different people.
So, to move away from experimentation and consistently create value, we need to help people:
identify the patterns in their work (their workflows)
figure out which of these workflows deliver the most value, and
define the work.
And once this is done, how they use AI or automation becomes glaringly obvious. And for many, that’s often enough to figure out what they need to do next.
For everyone else, this gives them a platform upon which they can start to experiment: “I’ve identified my repetitive, high-friction tasks - now I can figure out how AI can help”.
Thankfully, there are simple ways to help people build workflow capability. Because people aren’t lacking tech skills - they’re lacking awareness of what’s needed, an opportunity to practise and simple tools to prepare.
Over the next few posts, I’ll share what this looks like in practice:
how to run workshops that give people the chance to practise
how to build simple tools that guide them through it, and
how to bring it all together across the organisation, so it spreads beyond the enthusiastic minority and becomes a cultural shift.
And if you’re an L&D professional who wants to master workflow improvement and become the person responsible for unlocking the value of AI in your organisation, join my email list today.
Next article: Part 2: How to Design an Enterprise AI Workflow Capability Program.