Closing the Gap Between AI Investment and AI Impact with Max Haining, 100 School
Max Haining, founder of 100 School believes companies are failing because they've given people access without giving them a structured way to...
Watch videoIn this episode, we talk to Aaron Asadi, CEO of Enterprise Nation. Aaron draws on his background running massive digital divisions at Future and Saga to discuss the unique challenges facing the UK’s 5.8 million small businesses. He cuts through the “magic beans” marketing of AI to explain why the real opportunity lies in redesigning workflows around talent, not just using LLMs for superficial tasks. What you will learn:
The article – Why most small businesses aren’t getting past the basics with AI
John Emmerson: Hi, I’m here with Aaron Asadi, CEO of Enterprise Nation. Before Enterprise Nation, you ran large digital businesses at scale — hundreds of millions of users at Future, and you launched Saga’s digital division. Now you’re running a very large network where most members have ten or fewer employees. What are you seeing as the biggest differences, and what surprised you most about what small businesses actually need versus those enterprise-level organisations?
Aaron Asadi: I don’t know that I was surprised, I suppose. Growing up, my dad was a small business owner, not a very successful one. I’ve always known that small business is hard. In a small business, there’s generally a person who’s very good at something and there are gaps. How you fill those gaps is really, really difficult for a small business owner. In a corporate, those gaps are filled by other departments. But where I’ve been more surprised is how things haven’t really changed in terms of priorities over the years for small businesses. People are still struggling to access finance, still struggling to access new skills. And digital is often used as a synonym for easier, but I don’t think that’s the case.
Aaron Asadi: Coming to Enterprise Nation, what was really surprising was what sat underneath those top-level categories. Only 15 per cent of small business owners in the UK are women — and that’s in decline. That’s obviously not a reflection of ambition or talent. It’s clearly a demonstration of something that is systemic and going wrong, and that’s for half the population. It’s the same for people from ethnic minority backgrounds — they’re only 11 per cent of the small business population, which is far below their share of the working age population. And on top of that, both those groups find it much harder to access finance. Beneath all of that, the thing I’ve been really taken with is some of the data we see around loneliness and isolation. Small business owners are really difficult. A lot of the time they’re alone or they feel alone. You can see the evidence in how that impacts mental health. So the real issues are still the same, but I’ve been taken aback by some of the more human and systemic challenges we’re facing.
John Emmerson: What are you at Enterprise Nation doing to help those groups?
Aaron Asadi: We’ve run a number of programmes over the years, which is how we’ve been able to amass a larger percentage of our membership as female or from ethnic minorities. We’ve run programmes with Uber supporting Black-owned businesses — a big push on Black-owned restaurants for four years. We run various female founder campaigns and we’re just setting up a new female founder micro-community within ours because we want them to feel connected and feel stronger as a group. But it’s really hard, because we’re not a bank, we’re not a lender. All we feel we can do is represent their needs and be a voice for them. Affecting change is very difficult.
John Emmerson: You have a small business barometer which tracks what’s happening across UK businesses. Where are the bright spots?
Aaron Asadi: There’s not many, and I don’t want to patronise folk by waving a flag and saying we’ve got this, because it’s uniformly hard. What we’re blessed with in this country is that we’re a really well-educated nation with a history of great innovation, tremendous resilience, and great passion for ideas. We’ve always been at our strongest when we’re more diverse and more willing to embrace the diversity of ideas. So you can see the foundations are there for us to be a very successful entrepreneurial society.
Aaron Asadi: I was meeting recently with a company of about ten or eleven people — brilliant scientists who’ve developed and earned a patent for an alternative to antibiotics using photonics to attack bacteria. Classic UK innovation. Wonderful passion, great ideas. Struggling with funding. Struggling with marketing. And likely to go elsewhere outside this country to really make a difference. That’s emblematic of where we are. I see a great deal of talent and great ideas. It’s moving onto that next stage where it is hard to grow. The economy is growing at whatever the latest figures are — 0.1 per cent. Brexit has been a disaster. Costs continue to go up. But when you look at 5.8 million small businesses driving 52 per cent of the UK’s business turnover, the positive for me is less about what’s happening, which is tough, and more about what the potential is. If we can just make small business management easier by ten per cent across all those businesses, you’ve grown the business turnover of this country by five per cent. That’s the positive. The potential is the positive.
John Emmerson: Your data shows 79 per cent of small businesses are using AI in some context, and 84 per cent of your founders are ready to engage with it. But nearly half say they lack internal expertise. What needs to happen to close that gap?
Aaron Asadi: I think some of that gap is healthy right now, in so much as founders are experimenting, not implementing. And I think that’s been okay while it’s been uncertain which tech to back. Only two years ago, no one had really heard of Anthropic or OpenAI’s models. So it’s quite hard for people to really invest while there’s no sense of that stabilising. I think the experimentation has been a really good sign. The hesitancy to implement is partly down to trust in the tech. If I hinge my business on Anthropic today, is it going to be there? These are the doubts people have.
Aaron Asadi: We are starting to see that shift from experimentation to implementation. We’re seeing AI creep into people’s workflows, but those workflows aren’t being designed around AI. From what we can see, it’s still being used effectively as a surface layer tool. People are running the business as they normally would — same spreadsheets, same systems — and consulting an LLM to check something or get their copy checked. There’s a misunderstanding going on about what AI actually means. A lot of people just mean LLM right now. I think it’s about making it clear that there are ways to redesign your entire workflow around AI, as opposed to something as superficial as saying: I use ChatGPT every day because I ask it questions. That’s a very different thing.
John Emmerson: You’ve said previously there’s no greater opportunity for small businesses to grow than with AI. Do you still believe that?
Aaron Asadi: Yes, a hundred per cent. I feel very strongly that these tools are amazing when used properly. A small business owner probably got into business because they’re really great at something. But running a successful business requires a broad range of skills. I’ll go back to the example of my dad — he could fix anything, genuinely brilliant at that. But he wasn’t an accountant. English was not his first language. He wasn’t a marketing guy. The business failed. If I could go back to the nineties and give him even a ChatGPT, his world would have been transformed. He couldn’t afford a team around him to fill in those gaps. So this tool comes in and fills them. Maybe in every area it’s not the very best — maybe it’s 70 per cent of the best copywriter you’ve ever seen — but 70 per cent is infinitely more than zero per cent. It’s potentially a great enabler. Everyone suddenly has their own Jarvis system, just with them all the time.
Aaron Asadi: AI has that great opportunity to become everything else, so that the stuff you’re really good at and really passionate about, you can spend time on it and get better at it. The opportunity for me with AI isn’t so much that you’ll get the best possible marketing copy or the best accounts — it’s that you won’t have to worry about that stuff so that you can do the very best at the thing you yourself are particularly good at. You want artists to be artists. You want mechanics to be mechanics. You want craftspeople to spend their time making craft. You don’t want them doing stuff they didn’t get into business to do. That’s the opportunity. It’s an accelerator for the stuff they’re not supposed to be doing.
John Emmerson: A lot of small businesses using AI are using it to pump out content — social posts, messages. And a lot of it is clearly AI-generated. Is that helping them or harming them?
Aaron Asadi: Most of my career has been in content — twenty years in publishing. If you go to Threads, anything AI gets called AI slop. That’s become an automatic position, to the point where some stuff isn’t even AI but people don’t like it, so they call it that anyway. We’re going to see a growth in volume without doubt. But in terms of impact on content, I think it depends on context. Media organisations have been writing algorithmically for years. A press release is written in a certain format. A news story is written in a certain format. I used to run large consumer websites and we were very specific around writing to appease Google’s mysterious search algorithm. AI can speed all that up. But AI needs the data. And the data for really valuable content — AI can’t listen to Taylor Swift, it can’t go to a war zone, it can’t test a toaster. The human experience really informs great content.
Aaron Asadi: When it comes to the more artistic side, my views are perhaps atypical. For me the question is: does AI help the expression of an original idea, of an important idea, and connect people — which is the real purpose of art? I’m definitely not in the camp of everything AI is AI slop, not at all. So long as it’s doing something valuable and coming from a very real place where someone has had a genuine idea and wants to affect something or communicate something — AI can be a paintbrush just as much as a paintbrush can be. My son is autistic and has always been particularly unable to hold a pencil firmly, which effectively disables him from drawing the character he wants to draw or hand-writing the story he wants to write. Does that mean his ideas, thoughts and feelings should not be expressed? If you go down the route of everything AI is AI slop, that’s the logical extension. To dismiss something purely because it was made with AI is narrow-minded.
John Emmerson: Are you currently working with anyone specifically on AI education and adoption?
Aaron Asadi: Yes. More recently in the AI space specifically, we’ve just started running some programmes with OpenAI, and we just finished supporting Google on its Google AI Works for Business series of workshops, which we think will continue this year. We help Google with all its Google Digital Garages — about 50-plus a year. A subset of that now is the AI Works for Business programme. It’s really important that we’re working alongside Google and OpenAI because these organisations are really going to be making a big impact in AI. We want to be close to them and have effective partnerships alongside the organisations we trust. And the people I’ve met in person at those organisations are really passionate about this. They genuinely want to see change in this area, which makes us feel good about the partnerships.
John Emmerson: What is the most common mistake you see small businesses making with AI right now?
Aaron Asadi: I’ve mentioned a couple. I think the most common one is thinking that AI equals LLM. Even a fairly medium-sized business might be using Xero, their own spreadsheet on Dropbox, some Microsoft gear, and something offline that’s been their finance system for ten years. This idea that you can just bring in AI and improve a particular function — some of the feedback you’ll get is: well, ChatGPT was no good at all. And that’s because ChatGPT isn’t going to solve this very unique combination of platforms and your own operating system. So AI is not just an LLM. It’s understanding that AI is lots of different things. And closely connected to that: AI solutions are not going to be simple. When they’re effective, they have to be bespoke, have to be localised. You’ve got to consider your entire existing operating system to get the benefit out of it. It’s a serious project. In the same way that migrating from one CRM to another is a big project — steering meetings, trade-offs, arguments, training, it takes ages — AI is no different and should be no different. We’ve got to be clear about what the solution is. It might be different tools, it might be an agent developed in your own engineering team. AI solutions, when they’re effective, are bespoke. That’s the mistake I see most.
Enterprise Nation’s research points to a potential five per cent increase in UK business turnover if small businesses — which drive 52 per cent of the country’s business turnover — can improve their efficiency by just ten per cent through AI. Aaron Asadi describes this as the ‘potential dividend’: not a reflection of what is currently happening in the economy, which he characterises as very difficult, but of what becomes achievable if AI is adopted properly at the small business level.
Aaron Asadi identifies two main reasons. The first is uncertainty about which technology to back: with models changing rapidly and new providers emerging constantly, many founders are reluctant to commit their business processes to a platform that may look very different in twelve months. The second is a misunderstanding of what AI actually is. Most small businesses currently equate AI with LLMs and use it as a surface-level tool — checking copy, answering questions — rather than redesigning their workflows around it. Both factors are healthy in the short term but will become a constraint as more sophisticated competitors move from experimentation to implementation.
Aaron Asadi’s argument is that AI is the first technology capable of filling the skill gaps that prevent talented specialists from focusing on what they are actually brilliant at. A founder who is exceptional at their craft but not an accountant, marketer or administrator can use AI to reach a functional level — perhaps 70 per cent of the best professional in each of those areas — rather than zero. The goal is not to replace expertise but to free founders from the work they did not start a business to do, and give them back time to develop the work they did.
Aaron Asadi’s view is nuanced. He rejects the blanket dismissal of AI-generated content as ‘slop’, arguing that what matters is whether the content expresses a genuine, original idea and connects people — not whether the tool used to produce it was a human hand or an AI. He draws a distinction between formulaic content, which AI can produce efficiently and which the industry has been writing algorithmically for years anyway, and content with genuine human insight — war reporting, first-person experience, creative vision — where the human element remains irreplaceable. He is candid that volume will increase, and that authenticity and originality become more valuable as a result.
The most common mistake, according to Aaron Asadi, is treating AI as synonymous with LLMs and assuming a simple solution exists for complex operating environments. Most small businesses have a unique combination of systems, platforms and workflows built up over years, and expecting a single AI tool to work across all of them without integration work is unrealistic. Effective AI implementation is a serious engineering project, equivalent in scale and complexity to migrating a CRM — it requires clear definition of the problem, consideration of the full existing operating system, and often bespoke or localised solutions rather than off-the-shelf ones.
Aaron Asadi notes that only 15 per cent of UK small business owners are women — a figure that is in decline — and only 11 per cent are from ethnic minority backgrounds, far below their share of the working-age population. Both groups face compounding disadvantage: not only are they underrepresented as founders, but they also find it significantly harder to access finance, making their businesses less likely to survive. Enterprise Nation works to address this through targeted programmes with partners including Uber, and through community-building initiatives for female founders, but Aaron is direct that without systemic change in finance access, the impact of these programmes is limited.
Enterprise Nation is currently running programmes in partnership with OpenAI and Google, including Google’s AI Works for Business series of workshops and Google Digital Garages. The organisation’s role is to help design and deliver programmes that match its members’ needs and values, using the reach and technical expertise of large AI companies while ensuring the education is grounded in the practical realities of running a small business. Aaron Asadi emphasises that these partnerships work because the people driving them within those organisations are genuinely passionate about improving AI literacy at the SME level, not simply using Enterprise Nation as a distribution channel.
Max Haining, founder of 100 School believes companies are failing because they've given people access without giving them a structured way to...
Watch video
"For the last two years, organisations have been chucking mud at the wall," says Max Haining. "Now lots of...
Read more
Not long after ChatGPT fell on us back in 2023, my colleagues and I started scoping AI training courses. What...
Read moreGet ahead with the most actionable insights, playbooks and real-world AI use cases you can adopt right now, in your inbox every week