78% of businesses are using AI.Only 9% have made it work. Here's what changed the conversation.
By Hannah McKernon, Heads of Digital Marketing at Maybe*
On Tuesday, 24 March, our CEO and Founder, Polly Barnfield OBE, took the stage at Google's ‘Empowering Entrepreneurs in the Age of AI’ event in London, alongside Innovate UK.
The room was full of founders. People are building real businesses, trying to work out where AI fits, what it actually does, and whether the gap they feel between themselves and the big tech narrative is real or imagined.
It is real. And it matters. Here's what the conversation surfaced.
The gap nobody's talking about, honestly
The headline AI number is impressive. Most businesses have tried it. A significant proportion uses it daily. But our research, drawn from 1,500+ interviews with founders, marketers and business leaders, tells a more complicated story.
78% of businesses are using AI in some form
9% have genuinely integrated it into how they operate
18% of a marketing team's week disappears managing disconnected AI tools before real work starts
That gap between adoption and integration is where most AI investment quietly disappears. And it's where the conversation at Google landed hardest.
“The founders winning with AI aren’t using more tools. They’re doing less with more clarity.”
What founders in the room were actually asking
Three questions kept surfacing, in different forms, across the whole event.
QUESTION 1
How do I prove ROI from AI before anyone takes it seriously internally?
The honest answer: you can't prove it if you didn't measure it from the start. The businesses getting real returns set a baseline on day one, even a rough one. They track one thing. They build from there. No baseline means no proof, and no proof means the next budget conversation goes badly.
QUESTION 2
What should I actually start with if I'm overwhelmed by the options?
Not a strategy document. Not a tool audit. One task. Pick the thing in your business that takes too long, gets done inconsistently, or falls through the cracks every week. Make that work properly. Then add another.
That's the adoption pattern that compounds. Everything else is procrastination dressed as planning.
QUESTION 3
Is the access gap real, or is it just a matter of trying harder?
It's real. The tools, networks, investment routes, and visibility that make AI adoption easier are still unevenly distributed, concentrated in major hubs, certain sectors, and businesses with existing technical resources. That's not an excuse. But it is a system problem, not a personal one. Naming it matters. It's the first step to fixing it.
What the research already told us and what the room confirmed
Maybe* has spent the last year talking to over 1,500 businesses about AI. Not the polished case studies. The real experiences. What's working behind the scenes, what's failing quietly, and what separates the 9% who've genuinely integrated AI from the 91% who are still experimenting at the edges.
The patterns from that research showed up in that room on Tuesday. The same pain points. The same hesitations. The same breakthrough moments.
Four things stood out:
Integration beats adoption. Having AI tools is not the same as having AI working. The businesses making progress have embedded it into existing workflows within the tools their teams already use, not bolted it on as an add-on.
Disconnection is expensive. Teams lose almost a full day per week managing AI tools that don't talk to each other. That's not a technology problem. It's a connection problem.
Trust is built by finished work, not by promises. The businesses with the highest AI trust aren't the ones with the most sophisticated tools. They're the ones where work visibly completes itself, where an instruction goes in, and a finished outcome comes out.
Starting small is not the same as thinking small. The businesses that scale AI fastest start with one task, prove it, and build from there. The ones that try to transform everything at once usually transform nothing.
“Everyone is trying AI. Very few see finished work appear.”
What happens after an event like this
The conversations in the room matter. But what matters more is what happens after people leave.
Most founders walk away from AI events with a longer reading list and a vague sense that they should be doing more. We wanted Tuesday to be different. We wanted people to leave with one concrete next step, not a strategy, not a tool list, but a task.
If you were in the room: what's the one thing in your business that Maybe* could handle this week? That's the question worth sitting with.
If you weren't in the room: the research is where to start. Everything we covered on stage, the data, the patterns, the honest picture of where AI is working and where it isn't, is captured in The Big AI Secret.
Read the research
1,500+ interviews. 12 chapters. Every finding that shaped how Maybe* is built.→ Read The Big AI Secret
Start with one task
Tell Maybe* what needs doing. It completes the work.
