It's easy to dismiss AI as hype. The numbers say otherwise: this is one of the fastest technology shifts business has seen, and it has already left the experimental stage. What follows is a short look at where adoption stands, what the economics suggest, and why most companies still aren't seeing the results the headlines promise.
Adoption is already here
According to McKinsey's State of AI 2025, 88% of companies now use AI in at least one function – up from 78% the year before. Two in three use it in multiple functions. And 62% are already experimenting with AI agents.
The pace matters as much as the level. A ten-point jump in a single year is fast for any technology, and the spread into several functions tells you this is no longer confined to a curious team in one corner of the business. It is becoming part of how ordinary work gets done, from support and sales to finance and operations.
In other words: the question is no longer whether your competitors use AI, but how well.
The economic potential
PwC estimates AI could add $15.7 trillion to the global economy by 2030 – roughly 14% of global GDP. McKinsey calculates that generative AI alone could contribute $2.6–4.4 trillion a year.
The figures are staggering, but the point is simple: a large share of future productivity will be built on this. Numbers at that scale are abstract by nature, and no single company will feel them directly. What they describe is a direction of travel. The value shows up one workflow, one team, one freed-up hour at a time, and it accrues to the companies that learn how to capture it rather than the ones that simply own the tools.
But most don't get the value
Here's where it gets interesting. Despite high adoption, only around 6% of companies report that AI delivers significant impact on the bottom line. A third have begun to scale in earnest – the rest get stuck in the pilot stage.
The gap between using AI and getting value from it is the real challenge right now. Adoption is easy to count; impact is harder to produce. A tool can be in everyone's hands and still change nothing that shows up in the results.
Why the value gap exists
So why does the gap persist when the technology clearly works? In our experience it rarely comes down to the models themselves. It comes down to everything around them – the data, the ownership, and the path from a promising test to something a business can rely on.
- Pilots that never reach production – a demo proves the idea, then stalls because no one owns the harder work of making it reliable, monitored and safe enough to run every day.
- Scattered data – the information a workflow needs sits across different systems, formats and inboxes, so the AI never sees the full picture and its output can't be trusted.
- Unclear ownership – when a project belongs to everyone and no one, it drifts. No single person is accountable for the result or for moving it from idea into daily use.
- No measurement – without a baseline and a clear definition of success, no one can say whether the tool actually saved time or money, so it quietly fades away.
Each of these is a reason a pilot stays a pilot. Closing the gap is mostly the unglamorous work of taking one workflow from promising demo to dependable production – giving it an owner, clean inputs, and a number that proves it worked.
Two companies, same tools
Picture two companies with access to exactly the same AI tools. The first hands out licences, encourages everyone to try the chatbot, and points to the rollout as proof it is keeping up. People use it for the odd email or summary, but nothing changes in how the business actually runs. The activity is real; the value is not.
The second picks one workflow that genuinely hurts – say, the inbound customer questions that keep pulling staff away from harder work. They give it an owner, connect it to the right data, let it handle the repetitive cases, and measure response times before and after. It is narrower and far less impressive on a slide. But it frees real hours, and those hours fund the next workflow. The difference between the two isn't the technology. It is the discipline of finishing one thing well before moving to the next.
Closing the gap
This is exactly the gap we help companies close. We are less interested in adding another tool than in taking one workflow that matters and carrying it all the way to real, measured value. If you recognise your own company in either of the two above, we'd be glad to talk – reach out to Nodal and we'll start with the single workflow worth getting right first.
