AI Without a Business Problem 

Pearce still runs into organisations that start with the assumption they need an AI strategy, instead of asking what problem AI is actually suited to solve. 

"I have seen lots of organisations over the years start with 'we need an AI plan, go and figure it out,' rather than try to solve a business challenge that AI is suited to." 

He recalled one large global organisation determined to prove its AI success. When pushed for an example of value delivered, the only metric they had was employee adoption of Microsoft Copilot. 

"They had no use cases they had solved. They had no examples where they had saved time, done something innovative or positively impacted customers. Their only data point was that people have used it. People had used it, but no one could articulate any business value." 

Organisations are getting more strategic, Pearce said, but "we need to do something with AI now" still shows up more than it should. 

 
A Simple Framework for Finding AI Opportunities 

Forget the technology first. Pearce works from two questions instead: 

1. Where can we remove toil? 

Every organisation has repetitive, manual work that eats time without creating value: moving data between systems, filling out forms, routine admin. Necessary, but it doesn't need a human doing it. Cutting this toil is usually the fastest way to show measurable AI value. 

2. What can we now do that we couldn’t do before? 

Once that time is freed up, new opportunities open. AI shouldn't just automate what already exists, Pearce says. It should enable what wasn't practical before. 

His own podcast is the example. He liked recording episodes and talking to guests, but building promotional content for every platform was repetitive work that capped his reach. So, he built an AI agent that pulls the best moments from each episode, cuts short clips, writes platform-specific copy for TikTok, Instagram, Facebook, X and LinkedIn, and schedules it automatically. 

"I used to only post on LinkedIn. But now with AI I have reduced the toil and done something new, leading to more outreach." 

 
From Adoption to Impact 

Pearce's experience mirrors where enterprise AI is heading. The question is moving from where AI can be deployed to where it actually solves something. Start with the challenge, not the tool, and you get a clearer path to results, whether that's cutting repetitive work, improving the customer experience, or unlocking a way of working that wasn't possible before. 

To hear more from Ben Pearce, including a highly relevant conversation he had with Matt Quinn on overcoming resistance to disruptive technology, listen to this episode of Tech World Human Skills.