A study released this week by PwC dropped a number that should worry every small business owner in Europe: 74% of AI's economic gains are being captured by just 20% of companies. The other 80%? They're spending money on AI tools, running pilots, attending webinars — and seeing almost nothing in return.
That's not a small gap. That's a chasm.
But here's what the headlines miss: the companies winning at AI aren't winning because they spend more. They're winning because they use AI differently. And that's actually good news for small businesses willing to rethink their approach.
The study behind the numbers
PwC's 2026 Global AI Performance Study surveyed 1,217 senior executives across 25 sectors worldwide. The top-performing 20% are generating 7.2 times more AI-driven revenue and efficiency gains than the average competitor. Meanwhile, 56% of all companies surveyed said they've seen no significant financial benefit from AI at all.
Read that again: more than half of businesses investing in AI aren't getting measurable results.
For small businesses, the picture is even more nuanced. Adoption is at an all-time high — 68% of small businesses now use AI regularly, up from 48% in 2024. But only 29% of SMEs have reached what researchers call the "scaling" phase, compared to nearly 50% of large enterprises. Small businesses are quick to start, but they hit a wall when it comes to turning experiments into real business impact.
Why big companies pull ahead (and it's not budget)
The instinct is to blame resources. Of course a company with a dedicated AI team and a seven-figure budget gets more out of AI. But PwC's research points to something more interesting.
The top performers aren't just automating existing tasks faster. They're using AI to rethink how their business works — finding new revenue streams, entering adjacent markets, and redesigning customer experiences. PwC calls this "industry convergence," and it's the single strongest factor separating leaders from laggards.
A bakery that uses AI to answer emails faster is automating. A bakery that uses AI to predict demand, personalise loyalty offers, and show up in AI-powered search results when someone asks "best sourdough near me" is transforming. Same technology, completely different outcomes.
The second factor? The 20% don't treat AI as a standalone tool. They embed it into their core workflows. Their website talks to their CRM, which feeds their marketing, which adjusts in real time. For most small businesses, AI still lives in isolated pockets — a chatbot here, a content tool there — with no connection between them.
The European angle: 13.5% and falling behind
Here in Europe, the numbers are even more stark. The European Commission recently acknowledged that only 13.5% of EU businesses are using AI at all. That's across all business sizes — for micro-businesses and solopreneurs, it's likely even lower.
The EU is responding. Its new "Apply AI" initiative, backed by €1 billion, aims to accelerate AI adoption across strategic sectors. The Commission has deployed 19 AI factories across member states and is launching a network of Digital Innovation Hubs to help smaller companies access AI tools and expertise. Article 57 of the AI Act requires every member state to establish at least one AI regulatory sandbox by August 2026, giving startups and SMEs a supervised environment to test AI systems without the full weight of compliance.
These are real, funded support structures. But most small business owners I've spoken with haven't heard of any of them. The gap isn't just technological — it's informational.
What the 20% do that the 80% don't
PwC's study and broader research on SMB AI adoption point to four things that separate companies getting real value from AI from those stuck in pilot mode:
They start with a business problem, not a tool. The 80% typically adopt AI because it seems like they should — they sign up for ChatGPT, try a few prompts, and move on. The 20% start by identifying a specific bottleneck: "We're losing leads because our website doesn't answer questions after hours" or "We spend 12 hours a week on invoicing." Then they find AI that fixes that exact problem.
They connect their systems. An AI chatbot on a website that can't access your product catalogue, pricing, or booking system is a novelty. One that can actually answer "Do you have this in stock?" or "Can I book for Thursday?" — that's a business asset. The value comes from integration, not from any single tool.
They measure outcomes, not activity. Using AI to write 50 social media posts a week means nothing if those posts don't bring in customers. The top performers set concrete metrics before deploying AI: response time, lead conversion, hours saved on admin, customer satisfaction scores. If the numbers don't move, they change their approach.
They make their business visible to AI. This one is underrated. As more customers use AI assistants, voice search, and AI-powered recommendation engines to find businesses, the companies that show up are the ones with structured data, clean websites, and content that AI systems can actually read. A gorgeous website built entirely in JavaScript with no server-rendered content might look great to humans — but it's invisible to GPTBot, ClaudeBot, and every other AI crawler.
Three things you can do this month
You don't need a six-figure budget or a data science team. Here's where to start:
1. Audit your AI readiness. Before buying any tool, understand where you stand. Can AI crawlers read your website? Is your business data structured? Do you have the basics — a proper sitemap, schema markup, fast load times? Tools like Cresly's free AI Readiness Scan can give you a clear picture in minutes, showing exactly where your online presence falls short for AI visibility.
2. Pick one workflow and automate it end-to-end. Don't spread AI across five different tasks. Choose the one that costs you the most time or money — customer support, appointment booking, lead follow-up, invoicing — and automate the entire workflow from trigger to resolution. A single well-integrated automation beats ten disconnected AI experiments.
3. Make your website AI-ready. This is the lowest-hanging fruit and the most overlooked. Add structured data (JSON-LD) to your key pages. Make sure your content is server-rendered, not hidden behind JavaScript. Create an llms.txt file. Allow AI crawlers in your robots.txt. These are small technical changes that dramatically improve how AI systems discover and recommend your business.
The gap is closable
The PwC study paints a stark picture, but it also contains a hidden opportunity. The 74/20 split exists because most companies — large and small — are still figuring out how to use AI effectively. The playbook isn't locked behind a paywall or reserved for enterprises with deep pockets. It's about making smart, focused choices about where and how you deploy AI.
For European small businesses, the timing is actually favourable. The EU's regulatory sandboxes are opening up. Digital Innovation Hubs are being funded. And the tools available to a two-person agency in Amsterdam or a family restaurant in Lyon are genuinely comparable to what enterprises were using just two years ago.
The question isn't whether your business can afford to adopt AI. It's whether you can afford to be in the 80% that doesn't.