Everywhere you look, somebody is talking about AI. Prompts, copilots, chatbots, and digital assistants are appearing in business plans and conversations.

But here’s the question that really matters.

What do you want the technology to achieve?

The goal isn’t simply to buy more technology. It’s to give your people more time to think, solve problems, support customers, and grow the business.

Why starting with AI can be the wrong move

One of the easiest mistakes to make is starting with the technology rather than the problem.

The questions quickly become:

• How could we use AI?
• Should we buy an AI assistant?
• Which AI platform should we choose?

There’s a better question:

Think about the everyday work that quietly eats into the week.

• Reports that are recreated regularly.
• Emails that are sent again and again.
• Spreadsheets that someone updates manually.
• Information copied between systems.
• Folders checked just in case something has changed.
• Routine notifications sent by hand.

Putting clever technology on top of an inefficient process doesn’t necessarily fix the process.

In fact, it may simply help you do the wrong thing faster.

The four-part automation test

You don’t need a complicated assessment to find good places to start.

Look at a task and ask four questions.

If the answer is yes to all four, automation is probably a good fit.

Small tasks create a surprisingly big hidden cost

Half an hour doesn’t sound like much.

But imagine one repetitive activity takes just 30 minutes every working day.

Thirty minutes a day, across five days a week and 48 working weeks, becomes 120 hours a year.

That’s three working weeks spent on one recurring task.

And many organisations don’t have just one of these activities. They have dozens.

The biggest cost isn’t simply the time spent doing the work.

It’s what your people could have been doing instead.

A better order: stop, simplify, automate, then add AI

It’s tempting to jump straight to step four.

But the first three steps help create the foundations that make AI far more useful.

So where does AI earn its place?

Once you’ve dealt with the predictable work, AI becomes much more interesting.

AI can be useful where information needs to be understood, interpreted, organised, or turned into a useful first draft.

• Summarising large documents.
• Producing a first draft of a communication.
• Categorising incoming requests.
• Identifying patterns in information.
• Helping a person analyse information.
• Turning existing information into a useful starting point.

UK businesses are adopting AI, but there’s more to success than buying an AI tool

The interest in AI is real, and adoption among UK businesses has been growing.

But there’s an important detail in the same research.

The Office for National Statistics described adoption as relatively shallow. Among businesses already using AI, the average number of AI technologies being used had risen only modestly.

Separate Department for Science, Innovation and Technology research found that one in six businesses used AI, but most businesses had no active plans to adopt it.

That gives us an important reminder.

Simply having access to AI isn’t the same as getting meaningful value from it.

Not everything should be automated

Getting automation right isn’t about removing people from the business.

It’s about removing repetitive administration so your people have more time for the work where human experience matters.

Technology should support your people, not take important judgement away from them.

What could this look like in practice?

Imagine your team has a shared document library.

Today, somebody might:

1. Check the folder.
2. Find a new file.
3. Email the relevant team.
4. Start the review.
5. Update a tracker.

With the routine work automated, the process could instead start when the file arrives.

1. The file arrives.
2. The right people are notified.
3. A review task is created.
4. The team is updated.
5. The tracker is updated.

There doesn’t need to be any AI involved in that part of the process.

Once the routine steps are working smoothly, however, AI could potentially help a person work with the document itself. For example, the person could use AI to create a first summary or help analyse information.

That’s the difference between using AI because it’s fashionable and using it because it’s useful.

Measure the result, not the technology

It’s easy to count technology.

How many tools have we bought?

How many workflows have we built?

How many licences do we have?

Those numbers don’t tell you whether working life has actually improved.

Instead, look for outcomes such as:

• Hours saved.
• Fewer errors.
• Faster response times.
• A better customer experience.
• More capacity for valuable work.

Don’t start with a huge project

The words “AI” and “automation” can make the challenge sound enormous.

It doesn’t have to be.

Start with one frustration.

Prove the value.

Learn from it.

Improve it.

Then find the next opportunity.

The goal isn’t AI. The goal is creating capacity.

AI is a powerful tool, but AI shouldn’t automatically be the starting point.

First, find the repetitive work.

Simplify it.

Automate the parts a system can reliably handle.

Then look at where AI can genuinely help your people think, analyse, communicate, and make better use of information.

Most importantly, keep a person involved when judgement really matters.