:Harvey: Power Users: The Skill You're Sharpening

A power user is someone who, when faced with a problem, instinctively asks: “How can I use an AI tool to solve this?” rather than defaulting to a manual or traditional approach, even if that takes more effort upfront.

When you first start using Harvey or similar tools, you don’t see immediate productivity gains. It follows more of a J-curve — you’re experimenting, validating outputs, and building trust. But once you reach that point of confidence, the process becomes much more efficient — you know how to structure the request and get to a usable result quickly.

The number I typically cite is a 300% increase in productivity. This largely comes from comparing the time it takes to manually review due diligence reports versus running a standardised prompt across them. What used to take an hour and a half to two hours can now be done in about 15 minutes.

Without a tool like Harvey, I’d only have time to review the reports themselves. There wouldn’t be enough time to connect the dots between what’s in the data room and what’s reflected in the diligence. So it’s not just about efficiency, it creates the opportunity to do a deeper level of analysis. It’s effectively due diligence on the due diligence.

I manage a team of eight. My role is to set direction, while they’re executing day to day. If we’re expected to grow 25% year over year, there are only so many hours available. You either compromise on quality, or you find better ways of working — and for us, that means adopting tools like Harvey.

There’s a range of adoption across the team. Some people immediately see the opportunity, others take longer. But when the impact becomes clear, adoption follows. I use usage tracking as a directional tool, setting expectations and encouraging people to engage. And in many cases, the people who were initially hesitant end up becoming some of the most active users.

I lean toward them being made. It helps if you’re already comfortable with new technology, but this is different — you can become highly effective just by learning how to communicate with the tool. That lowers the barrier significantly.

Motivation can come from different places. For some, it’s about staying ahead of the curve. For others, it’s about working more efficiently and maintaining balance. You don’t need to be technical to get value, you just need to be willing to engage.

The core skill you’re developing is prompt engineering, understanding how to structure inputs, define outputs, and use consistent templates. Then there’s a second layer: breaking complex tasks into smaller steps and chaining those prompts into a workflow.

My framework is simple: the benchmark isn’t perfect use of Harvey, it’s what your work would look like without it. If you’re improving on that baseline, you’re making progress. You don’t need to be at the top immediately. The goal is to keep improving incrementally.

I don’t know exactly what this work will look like in five years, but I’m confident it won’t be a group of people sitting in a room manually reviewing due diligence reports.