Collective Campus
Blogs

Essay

Why Your AI Program Isn't Delivering Business Outcomes

Steve Glaveski

A dense magenta and cyan collage on dark plum

When I was 19, I embarked upon my first solo trip.

It was my first time catching a plane on my own, and as my taxi rolled up to the airport, I found myself running late.

So I made like Flock Of Seagulls’ classic hit, and I ran.

I ran through the terminal, working up a sweat as I dodged and weaved through the crowd like I was Muhammad Ali dancing around George Foreman.

And it paid off. I made it to the gate in time.

Except… I was in the wrong terminal.

Most work is a waste of time

You might think, “cool story Steve, but what on Earth does this have to do with how AI is showing up in our organisation?”

A lot, actually.

Back in 2020, at the height of lockdowns, I released my last book Time Rich (Wiley), which was an ode to better ways of working.

I made the case that most of what most people do at large organisations is a waste of time.

Price’s Law backs this up. It posits that the square root of an organisation’s headcount creates about half the value. That means that in a 1,000-person company, 32 people create half the value.

This is further echoed by the Pareto Principle, or the 80/20 Principle, which finds that 4% of causes create about 64% of outcomes.

Essentially, there is a very long tail of people doing work that doesn’t amount to all that much.

A typical workday

Think about your typical workday inside a large organisation.

  • Consensus seeking.
  • Hyper-responsiveness.
  • Always on.
  • Constant chatter (and GIFs) in Microsoft Teams or Slack.
  • Process paralysis.
  • Meetings that you don’t need to attend.
  • Making reports nobody reads.
  • Polishing work well beyond the point of diminishing returns.
  • Tenure and loyalty being rewarded over contribution.
  • Constant task switching.
  • Poor prioritisation of workload.
  • Completing tasks because we’ve always done them, but nobody seems to know why.
  • A long tail of busywork that adds up to very little.

Faster at wasting time

And then we add AI to the mix.

So now, we’re faster, better, stronger… at wasting time.

It’s akin to me running for my life through the wrong airport terminal.

Which is why 95% of companies are yet to see a measurable ROI on their AI investment.

Redesign the work first

To get business outcomes from AI, organisations need to revisit their culture, their processes, and how work gets done.

If leaders are serious about measurable business outcomes beyond simple declarations from employees like “I feel more productive”, then the work should be re-designed before bolting AI on top of it.

We’re going through a radical transformation of what “knowledge work” looks like.

This requires leaders to re-think everything they know about how work gets done and how value gets created.

It requires mapping tasks and workflows across the organisation, and re-designing them with AI in mind.

The hard decisions

It requires addressing cultural shortcomings and making some very hard decisions.

Recently, Canva announced that it’s laid off 22% of its staff, and Microsoft offered a quarter of its staff redundancies.

This is to be expected. How you choose to handle that is another question, and beyond the scope of this article. But there are ethical and respectful ways of parting with talent in an increasingly uncertain environment.

Otherwise, you can keep measuring vanity metrics like “AI adoption”, and “AI certification badges earned”, just don’t expect it to translate into meaningful ROI.

Back to the airport

And in case you’re still wondering. I missed that flight and learned a very valuable lesson early into my travels around the globe. I’m pleased to say I haven’t missed a flight since because I was in the wrong terminal.

AI Design Sprints: from so-called AI adoption to business outcomes

See how AI Design Sprints redesign the work before AI goes on top of it.

Book a call