Roughly four in ten failed startups die for the same dull reason. CB Insights, which has analysed why startups collapse, found that running out of cash or failing to raise new money is the most common cause, cited in around 38% of cases. Not a weak product. Not a missing feature. Money. Matt Haycox, the entrepreneur and investor who has funded over £1 billion of UK business activity, has been making a version of this argument for years, and the current rush to bolt artificial intelligence onto everything has only sharpened it.
“Everyone wants to talk about AI like it’s a magic button,” Haycox says. “It isn’t. If your business can’t manage its money, AI just helps you go broke faster, with nicer dashboards.”
Across the more than 100 companies he has backed, Haycox has watched plenty of founders chase the new thing while the fundamentals quietly fell apart underneath them. The tool changes every few years. The reason businesses fail does not.
The shiny thing is rarely the problem
There is a familiar pattern. A founder feels behind, reads that competitors are using AI, and decides that is the gap. So the energy goes into tools and automations while the real issue, the one that actually threatens the business, sits untouched in the bank account.
Haycox sees it as a kind of expensive distraction. Working on the exciting thing feels like progress. It is easier to spend a week building an AI workflow than an afternoon staring at a cash flow forecast that tells you something you do not want to hear.
“The businesses that die rarely die because they didn’t have the latest software,” he says. “They die because nobody was watching the money closely enough, early enough. AI doesn’t change that. It just gives you something more interesting to look at while it happens.”
The lesson that cost him everything
Haycox does not talk about cash discipline from theory. In the 2008 financial crisis he lost almost everything and was declared bankrupt. The businesses he had built came apart when the credit dried up and the cash stopped moving, and he spent the years afterwards rebuilding from nothing.
The experience left him with a permanent respect for the one number most founders treat as an afterthought. He came out of it understanding that a business is only ever as safe as its cash position, and that everything else, the growth, the team, the ambition, sits on top of that single foundation.
“I didn’t go under in 2008 because I had a bad idea,” he says. “I went under because I didn’t respect cash enough. You only make that mistake once if you’re paying attention. It’s the most expensive lesson I ever had, and it’s the one I try to give people for free.”
What AI is actually good for
None of this means the technology is useless. Haycox is clear that AI can take real work off a small team, speed up the boring parts of a business and let a founder do more without hiring. Used well, it makes a healthy business more efficient, and there is a genuine edge in being early to that.
The catch is the word healthy. AI multiplies what is already there. Point it at a working model and it helps. Point it at a business that loses money on every sale and it simply helps you lose money on more sales, faster. The technology is an amplifier, and an amplifier does not care whether the signal underneath it is good or bad.
There is even a place for AI inside the finance function itself, in forecasting, in bookkeeping, in spotting patterns a busy founder would miss. But Haycox’s point holds. A tool that helps you read the numbers is only useful to a founder who has already decided the numbers matter.
The metrics that actually matter
When Haycox talks to a founder about money, he is not asking about revenue. Revenue is the number that flatters everyone. He wants to know the real margin on each sale, how much it costs to win a customer, and how long the business could survive if the next big payment arrived late.
Those are the numbers that decide whether a business lives. A company can be growing its turnover and quietly going broke at the same time, because every new sale loses a little money and volume just speeds up the bleeding. Knowing the difference is the whole game, and no piece of software will care about it on your behalf.
Cash is a discipline, not a feature
The reason cash gets ignored is that it is unglamorous. There is no launch, no demo, no announcement in managing your margins and chasing your invoices. It is just the quiet, repetitive work of knowing what is coming in, what is going out, and when.
Haycox argues this is exactly why it separates the businesses that last from the ones that do not. The founders who survive a downturn are usually the ones who knew their numbers cold before the downturn arrived. They were not smarter. They were just paying attention to the thing everyone else found boring.
Where founders should actually start
His advice to anyone tempted to spend their next month on an AI project is blunt. Get the cash position right first. Know your real margin on every product. Understand how long your business can run if the next payment is late. Then, with that foundation in place, let AI make the machine faster.
Founders who want help getting the fundamentals right before chasing the next tool can find more on Matt Haycox’s business mentoring approach.
The takeaway is not anti-technology. It is an order of operations. AI is a fine accelerator and a terrible foundation. Build the business that can manage its own money, and the tools will make it better. Skip that step, and no amount of software will save you from the one number that has always decided who stays in business.
