If AI made us work faster, why doesn't it feel that way?

AI can complete tasks in seconds that once took hours. So why aren't businesses seeing a bigger difference in how they work?

According to the NSW Government, 47% of NSW small businesses are using AI. Separate national research from COSBOA and Square suggests business owners using AI save around one to two hours a week. The same report suggests owners are putting that saved time back into running their businesses, rather than working fewer hours. So the time saved is real. It just isn't necessarily turning into time off.

That's a meaningful improvement. But considering how much faster some tasks have become, it raises an interesting question.

Why aren't we getting more time back?

Disconnected workflows

Consider a business using AI to respond to customer enquiries. The replies are faster, but the enquiries still arrive without the right information. Someone still needs to follow up, check availability, confirm pricing and decide what happens next.

The response takes less time, but everything surrounding it stays the same.

It's a hypothetical example, but it illustrates a distinction that often gets overlooked. A business doesn't operate as a collection of isolated tasks. Work moves between people, systems and decisions. Making one step faster doesn't necessarily improve the journey from beginning to end.

Moving bottlenecks

And sometimes, making one step faster simply reveals the next problem.

A proposal that previously took two hours to prepare might now take twenty minutes. But if it still spends three days waiting for approval, the overall process hasn't changed much. The bottleneck has moved, rather than disappeared.

More output, more work

There's another possibility worth considering. When something becomes easier to produce, we may start producing more of it.

More reports. More proposals. More emails. More content.

Each might take less time to create, but someone still needs to read, review, approve, respond to or maintain it.

The person generating the work experiences a productivity improvement. The person receiving it might experience the opposite.

That doesn't mean faster tools are inherently creating unnecessary work. But it does suggest that productivity isn't just about how much one person can produce. It's also about what that output requires from everyone else.

And much of that work depends on information that isn't particularly easy to access.

Missing information

Businesses often hold important knowledge across emails, spreadsheets, shared drives, disconnected systems and people's heads. AI might be able to prepare an excellent response, but only if it has the right information to work with.

If someone needs to spend fifteen minutes locating the latest pricing, checking a customer's history and confirming which version of a document is correct, the speed of generating the response becomes less significant.

This is where the conversation about AI can become a little misleading. We focus on what the technology can produce, while overlooking everything required to make that output useful.

Sometimes the problem isn't that a business needs a smarter tool. It's that its information is difficult to find, its processes are unclear or too much depends on one person knowing what happens next.

Measuring the wrong things

Saving time is relatively easy to understand. A task used to take thirty minutes. Now it takes ten.

But that doesn't tell us whether the business is operating better.

Are customers getting answers sooner? Are fewer things being missed? Is there less rework? Can someone take leave without becoming the subject of seventeen urgent messages?

These improvements are harder to capture in a simple time-saving figure, but they can make an enormous difference to how a business feels to work in.

It's also worth remembering that not every saved hour needs to become an hour of additional output. Sometimes the value is having fewer interruptions, more room to think or simply finishing the day without carrying quite so much into tomorrow.

Maybe the work needs rethinking, not just accelerating

None of this diminishes what AI can do. Faster drafting, analysis and automation can be genuinely useful, particularly for smaller businesses where time and resources are limited.

But technology works within the conditions we give it. If the underlying process is fragmented, the information is scattered or every decision depends on the same person, making individual tasks faster can only achieve so much.

The more interesting opportunity might be to examine the work itself. Remove unnecessary steps. Improve how information moves. Make decisions clearer. Connect the things that should already be connected.

Then use technology where it actually helps.

Because perhaps the biggest productivity opportunity isn't finding another way to complete the same work faster.

It's questioning how much of that work needed to exist in the first place.

← Back to insights