However, sometimes the better question is whether the systems you already have talk to each other properly.

When conducting our 2026 Complexity Survey, we asked 1,375 people working in UK organisations which technology investment would make the biggest difference over the next two years. Connecting existing systems came top, ahead of embedding AI into existing software and automating manual processes.

It seems businesses are as interested in getting more from what they already have as they are in buying something new.

Small frictions add up

Productivity conversations tend to focus on big things like investment, skills and emerging technology. Day-to-day niggles are usually more about switching between applications to finish one task, hunting for information stored in three places and re-keying data because two systems can’t pass it to each other.

Any one of these is a minor annoyance - across a team, or a whole business, the hours stack up.

68% of respondents think their organisation loses six or more hours a week to operational complexity, including manual processes, moving between systems and searching for information. Nearly 44% put it at more than ten hours.

Those are estimates, not measured losses, and they shouldn’t be read as a figure for UK business as a whole, but they show how much working time people feel goes on the mechanics of getting things done.

We also asked what hurts productivity most. Switching between systems was the top answer (17.4%), followed by lack of integration (14.5%), limited use of AI and automation (13.3%) and repetitive manual tasks (11.4%). Businesses are looking for new ways to use technology, but a lot of the frustration is still about how existing tools fit together.

Connecting systems tops the list

Looking two years ahead, 20.4% chose connecting existing systems as the investment that would matter most. Embedding AI into existing software got 18.6%, automating manual processes 15.7%, bespoke software 13.2%, better reporting and business intelligence 12.4% and replacing legacy software 11.6%.

None of this says new software or replacement projects are a waste. Some applications have hit their limit, no longer meet the business’s needs or are too hard to maintain, but the results do question the idea that progress always means bringing something new in.

Most technology estates grow piece by piece. A system gets added to meet a need, support growth or help one function, and each decision makes sense at the time. Years later there are lots of applications and a tangle of relationships between them.

Picture a business with a good CRM, a capable finance platform and software that runs daily operations, where staff still copy information between them by hand. No single system is at fault, they just don’t work together.

Better connections cut duplication, make information easier to find and let processes run more smoothly. They also get more from money you’ve already spent.

AI shows up the weak spots

Interest in AI adds another angle. Businesses want it to help people work faster, take over routine tasks and make better use of information. How well it does that depends partly on what it sits on top of.

In our research, 25.3% said their organisation’s push to invest in AI had uncovered systems problems, and 20.4% said it had uncovered data problems. Cost issues (19.6%) and people issues (14.3%) came up too.

If information is scattered across disconnected applications, or a process needs someone to step in at several points, AI won’t fix that by itself. It’s more likely to make the problem obvious.

That doesn’t mean holding off on AI until every system is connected and every process redesigned. But decisions about AI should take in the wider estate: how good and how accessible your data is, how applications share it, and the processes people rely on every day.

Sometimes the practical first step is integrating existing software or automating one task. Sometimes it’s modernising an application, or building bespoke software to bridge a gap between systems. Either way, understand the problem before choosing the solution.

Start with what people put up with

You don’t need to review every application you own. Start by finding where work is harder than it should be, and why.

That means asking how well a system supports the people using it, not just whether it technically works. Do people enter the same information twice, teams spend time chasing data that should be to hand, approvals or handovers slow down routine work, and have workarounds become so familiar that nobody questions them?

The answers show where a small, targeted fix could help. They also show which problems need new technology and which need better integration, automation or a change to the process.

Tie the business case to an outcome: less time on admin, easier access to information, quicker decisions, more time for work that needs people’s expertise. Decide how you’ll measure it too. Take a baseline before you add an integration or automate a manual step, so you can tell afterwards whether it worked.

Where this fits

UK productivity depends on far more than software, including investment, skills, infrastructure and the wider economy. Closing the gaps between your own systems won’t solve the national picture.

The Office for National Statistics’ August 2026 release put output per hour 0.7% higher in the second quarter than a year earlier, on its preferred administrative-data measure. The Labour Force Survey-based measure showed a 0.2% fall over the same period, so recent estimates are uncertain. Read the latest ONS productivity figures.

Against that backdrop, our survey points to somewhere practical to look: the everyday technology behind your operations. For some organisations that will mean new capabilities, for others it will mean closing gaps between systems, improving how information flows or removing manual steps that have become part of the working day.

So before adding another platform, ask a simpler question: could the technology you already have do more if it worked better together?