AI Doesn’t Fix How Your Business Operates

Why more technology, more tools and more speed don’t automatically create better business outcomes

There has never been more pressure on business leaders to adopt AI.

New tools are appearing almost daily. AI can analyse information, automate tasks, create content, support decision-making and increasingly perform work that previously required significant human effort.

The promise is compelling;

Move faster.
Work smarter.
Reduce costs.
Increase productivity.
Create new opportunities.
Grow.

But there is a question I think business leaders need to ask before adding another layer of technology to the business; What is actually getting in the way of the outcome we are trying to achieve? 

AI can make some things faster without necessarily making the business better, and I think this distinction is becoming increasingly important.

AI adoption is increasing. But where is the value?

Recent global research shows a significant gap between individual productivity gains and broader business outcomes. McKinsey's 2026 State of AI research found that 80% of respondents say AI has improved their individual productivity, yet only 37% report that AI has contributed positively to enterprise-level EBIT. Only 6% qualified as AI high performers.

That difference matters.

An employee becoming faster at completing a task isn't necessarily the same as the business becoming more productive, because if the next step in the workflow still creates a delay, information still needs to be checked manually, ownership is still unclear or a decision still sits with one person, the overall process may not have changed very much.

The technology has become faster. The business hasn't necessarily become better.

In New Zealand, PwC's 2026 CEO Survey found that 70% of business leaders believe their technology environment is ready for AI, yet 78% reported that AI had little or no impact on their organisation.

That gap caught my attention, because if the technology is ready, but the business isn't seeing meaningful impact, perhaps what is required is looking at how the business is operating around it.

Are we adding AI or changing how work gets done?

This is where I think the AI conversation needs to become more interesting.

Deloitte's 2026 State of AI research found that while access to AI across the workforce has grown significantly, only 30% of organisations are redesigning key processes around AI. Another 37% are using AI at a surface level, with little or no change to underlying processes. 

That distinction is important. There is a difference between adding AI to an existing process and rethinking how the process should work because AI now exists.

One adds technology, the other changes the way the business operates.

I've seen this pattern many times with technology, long before AI became such a dominant part of the conversation.

A business identifies a problem.
A solution is introduced.
The system is implemented.
People are trained.
The underlying workflow remains largely unchanged.
The technology is expected to solve the problem.

Sometimes it does, but sometimes the technology simply becomes another layer sitting on top of an operating environment that already contains friction, ultimately causing more friction.

AI can accelerate the wrong thing

Consider a workflow with unnecessary approvals, or a process that requires the same information to be entered into multiple systems, or information that sits with one person because the knowledge hasn't been sufficiently shared, or a team that spends hours checking and correcting information because the original data isn't consistent.

Adding AI to those environments might create efficiencies, but it doesn't automatically remove the reason those inefficiencies exist in the first place.

In some cases, it might simply make parts of the existing process faster, but speed doesn’t equal progress.

For example, you can automate a poor workflow, accelerate unnecessary work, produce information more quickly without improving the decision that follows and increase individual productivity while leaving the wider workflow unchanged. And this is where I think business leaders need to be careful.

This is where Hidden Friction™ matters

I've spent much of my career noticing that the problem being discussed isn't always the problem actually getting in the way.

A productivity problem can be a workflow problem.
A communication problem can be a visibility problem.
A people problem can be influenced by the operating environment.
A technology problem can be a design, adoption or configuration problem.

And what looks like a growth problem can sometimes be friction quietly accumulating underneath the business. Hidden Friction™ isn't necessarily failure or dysfunction, it’s the gap between how a company needs to operate and how it actually operates. It can be found in the small delays, repeated work, workarounds, unclear ownership, disconnected systems, unnecessary approvals and decisions that keep coming back.

These things can become so familiar that they stop being noticed, until the business tries to scale, then the friction that was manageable with a small team can become a significant constraint as the team grows.

The businesses getting more from AI are changing more than the technology

PwC's 2026 global AI Performance Study provides another useful perspective. Its research across 1,217 senior executives and 25 sectors found that the top 20% of organisations are capturing 74% of AI-driven returns. Those organisations are not simply deploying more AI tools. They are approximately twice as likely to redesign workflows around AI and two to three times more likely to use AI to identify growth opportunities and reinvent their business models.

That tells me something important. The difference isn't necessarily who has access to the technology because most businesses can access AI now. The difference is increasingly what they do with it. 

The strongest organisations aren't treating AI as another tool to add to the business, they're using it as an opportunity to reconsider how the business operates

AI needs to sit inside the operating model

This is the part of the conversation I think we need to have more often. AI shouldn't sit on top of the operating model as another layer, it’s a piece of technology that needs to work within an operating environment that is capable of using it effectively. That means;

Understanding the existing workflows.
Understanding where friction exists.
Understanding how decisions are made.
Understanding where information sits.
Understanding who owns what.
Understanding how systems connect.
Understanding what the client actually experiences.
Understanding what outcome the business is trying to improve.

Only then can leadership make better decisions about where AI should be applied, because sometimes the answer will be automation, but other times it will mean better data, a redesigned workflow, clearer ownership, a change in capability. And sometimes the right answer may be to do nothing with AI at all.

The technology should support the operating model, it shouldn't be expected to create one.

The question leaders should be asking

I don't believe the answer is to slow down the adoption of AI. There is enormous potential in the technology, and businesses that understand how to use it effectively will have significant opportunities to improve productivity, create new capabilities and find new ways to grow.

But speed of adoption shouldn't become the measure of progress, the better question is; What does our business need to change for AI to create meaningful value? That question moves the conversation away from tools and towards outcomes.

It asks us to understand before we automate and to look at the way the business actually operates before deciding what technology it needs.

AI can accelerate your business, but it can't decide what your business needs to accelerate. That remains a leadership responsibility and I think that is where the next stage of the AI conversation needs to begin.