What I Told MBA Students About Building an AI Startup

Yesterday, I spoke at Anglia Ruskin University to a group of MBA students who will be undertaking a mock consultancy project for MercuriDash. I had the opportunity to share parts of my story and journey over the last ten years — the journey of an unlikely entrepreneur. One who has had to figure things out mainly alone, with a lot of grit and on-the-job learning. Along with this, I spoke about how to build an AI startup for those keen to do that now.

When preparing for this talk, I found myself wondering: What can I actually teach a group of MBA students? After all, they probably know more business theory than I do. But then I realized: it’s not about theory. It's about sharing the battle lessons — the real, lived experiences that can't always be learned from a textbook.

The truth is, I didn’t do an MBA. In fact, I only studied Business Studies up to GCSE. I didn’t have any family members with successful businesses, nor did I have the opportunity to work with an entrepreneur. I don’t even remember ever being spoken to by one about it. I never knew that being an entrepreneur was even an option for me.

But here I am.

I began my entrepreneurial journey in 2014 with an idea to create a fashion marketplace. The business lasted four years, and while it didn’t succeed, I didn’t realize that I was laying the groundwork for what I’m doing today. Fast forward to now, and I’m running an AI-driven SaaS platform that aims to make fashion design and product development more efficient, cost-effective, and help brands make better product decisions.

Here’s what I shared with the MBA students:

1. Learn How to Learn

As an entrepreneur, this skill is absolutely vital. In the early days, you will likely be a one-person show, and you’ll need to be a generalist. You won’t know everything, and that’s okay. In fact, it’s expected. When I started my first business — a fashion marketplace — I had zero experience in business, and no prior experience in the fashion industry. But I just dove in. The key to making it work was being an active learner on the job.

You need to actively embrace the role of being a continuous learner in everything you do. Whether it’s reading, researching, or learning from others, you have to learn forever as an entrepreneur.

2. Resilience and Determination

Resilience is a tough one to teach because it’s something that’s either innate or developed through pushing yourself through difficult experiences. In pursuing challenging goals, you will inevitably encounter setbacks and failure. These obstacles don’t define you — they build the muscles for resilience.

I used to think failure meant the end of the road, but now I see it as part of the learning process. Statistically, the first business you start is unlikely to succeed — and when it doesn’t, it will likely be a public failure. The key is to own it. Share what you’ve learned from it, and move on. At least you tried. You now know more than someone who never tried and just trolls from the sidelines.

3. Know That Dots Only Connect Looking Backwards

When I started my first business, I was devastated when it didn’t work out. I believed I had failed at my dream. But now, I realize that if I hadn’t gone through that experience, I wouldn’t have gained the deep domain knowledge that I’ve since applied in building an AI business.

Similarly, my career path was anything but linear. I spent 17 years working as a lawyer, doing PR, and working in fashion retail. These seemingly unrelated roles all gave me unique insights that have proven valuable in running my startup.

As an entrepreneur, you might feel like you’re bouncing around, trying different things. But in the end, all these experiences add up. Whether it’s traveling, doing odd jobs, or learning about various industries, you’ll accumulate valuable life experience that will be crucial to your entrepreneurial journey.

So, let yourself live. The path ahead will likely be demanding and you won’t have much free time once you get going. But use this time wisely. Discover new things, explore, and get to know yourself.

My Tips For Building An AI Start-Up

1. You Don’t Need (Or Be) A Technical Founder

One of the most common misconceptions is that you need to be a coder to build an AI startup. I’m proof that this isn’t true. I can’t code. But over the years, I’ve managed tech teams, learned about both front-end and back-end development cycles, built AI models and have gained enough technical knowledge to supervise and critique effectively.

A technical co-founder can be a great enhancement to the team, but don’t let the absence of one hold you back. Don’t get stuck thinking you need one — learn what you can and stay deeply engaged with your product. It’s your company and your vision which needs to inform the user flows and AI model. Ultimately you might be grateful if you had to move forward without a technical mate as understanding every aspect of your product - including engaging with the technical build - will be of massive benefit to you later.

2. Understand Technology Trends

As a CEO, your role is managing resources, teams, and budgets. Familiarity with AI tools, no-code apps, and automation can save you time, money, and a lot of headaches.

We’re in a golden age where you can start a company on your own and build it before you need a massive team. This is where technology — especially AI — can be a game-changer. If you can use AI to boost efficiency and streamline processes, your business will have a significant advantage.

And if you have a team, encourage a culture of embracing AI within your team. There’s no shame in using AI to improve productivity, and rewarding efficiency is key - so make sure you communicate to them they should be using AI (responsibly of course and with checks and balances).

3. Domain Knowledge Is a Massive Edge

Anyone can learn to use AI models, tools, and APIs. But domain knowledge? That’s a different story. Having deep insights into an industry or specific problem is invaluable. Knowing the workflow, understanding the real friction points, and speaking the customer’s language are things that can’t be replicated by just anyone.

If there’s an industry or problem you have unique insight into, that’s your edge. If not, start by solving a problem that you’ve personally faced. If you know the problem intimately, that’s your unique advantage. Technology can be learned. Domain knowledge takes years to accumulate. And many founders trying to succeed at AI companies - and who have even raised millions of investment - will fail because they just don't get the industry and all of its weird ins and outs, peculiarities and nuances. You can't read that stuff in a book or even in a customer interview, you need to have spent time in it

4. AI Is a Tool — Not a Business Model

Don’t start with the idea of building an AI startup just for the sake of it. Instead, start with a painful problem, a broken workflow, an inefficiency or an opportunity to improve something in that world.

AI is an enabler, not a business model. Sure, it’s tempting to get caught up in the AI hype, but remember: you need to solve a real problem or create real psychological value. If you’re improving an existing system in a meaningful way, then the value will speak for itself.

Think about it this way: if people started using your AI solution, would they look back in five years and wonder how they ever lived without it? That's what social media did. It became so ingrained people cannot remember what they used to do before it existed.

5. Distribution > Technology

Many companies are using the same foundational models and technology, but what differentiates successful startups is distribution. Having access to customers, building trust, creating a brand, and fostering a community are the things that matter most. If you have a network or access to a community that's a great start. If you don’t already have a built-in network or community, think creatively about how to get your product to your target market.

Early-stage companies struggle with funding, so plan how you can spread your message cheaply. Do you have a personal brand? Can you leverage influencers or strategic partnerships? Does your message inspire and travel? And most importantly can you do this cheaply?

6. Consider Defensibility Continuously

In the past, companies could get away with slapping a fancy UI on top of a foundational AI model - think the AI assistant chatbot. Today, that’s no longer enough. You need to think about how you can defend your business against competitors. This goes beyond proprietary AI models, as many companies use foundational models in the start. Its not really about that alone - that's just one facet.

True defensibility is multi-faceted - it may include a couple of moats stacked on top of each other. Can you build workflows that are unique to your business? Can you create a system where data compounds over time, making your product harder to replace? Do you have a network advantage? The key is to create a uniquely, defensible position that others can’t easily replicate - even if they copied one aspect of your business. Think about which parts you would patent if you could - even if you never secure one its a useful framework to think about innovation and building on top of state of art.

7. Just Because You Can Build It Doesn’t Mean You Should

AI is an incredible tool for rapid prototyping, but just because you can build something, doesn’t mean you should. We’re going to see more and more unnecessary apps being built because it’s easy to do. Building something for fun is one thing, but building a business is something else entirely. Go back to first principles to decide whether this is something that should exist in the first place.

8. Build for Retention, Not Users

It’s easy to get caught up in the idea of gaining users, but retention is the true metric of success. Building a product that people love, trust, and rely on is what will ultimately determine your success. These days, there are too many apps and platforms out there fighting for attention. It's no longer enough to just acquire users — you need to create a product that people genuinely value and keep coming back to.

The days of free users being the main measure of success are long gone. Free is not free anyway - and attention is hard to get - so it’s far better to focus on creating a niche, highly engaged user base that truly loves your product. As competition grows, retention becomes more important than ever. The ultimate test is not how many users you have, but how many paying users you have — and how many of them continue to pay again and again.

To wrap up, here’s a quick recap of the key lessons and advice I left them with:

Start before you feel ready — Don’t wait for the perfect moment. Start now, even if it feels uncertain.

Learn relentlessly — Be a continuous learner. Read, listen, ask questions, and absorb everything you can.

Train resilience — Failure is part of the process. Use it as a learning tool, not as a roadblock.

Trust the non-linear path — Your journey might seem winding, but it’s the experiences you accumulate along the way that make you unique.

Gain deep domain knowledge — It’s your industry insights, not just your technology, that will give you the edge.

Validate real problems — Don’t build just because you can; build because there’s a real value in what you are building

Build for retention — User count is irrelevant if no-one is paying and continuing to pay. Retention is the metric that matters

Use technology wisely — Use AI and other tools to increase efficiency, but have checks and balances.

You don’t need permission — Your journey is yours to take. Don’t wait for others to tell you it’s possible.

You are the master of your destiny — Take ownership of your vision, and build the future you want to see.