Learning business by doing it

I’ve been building businesses since I was 18, and I taught myself every part of it. Stores first, then ads because the stores needed customers, then apps and automations because the work needed to run without me in every step.

How do you learn all that without being taught? Slowly, by doing it, and by paying for a lot of lessons. But there are ways to make those lessons cheaper, and I wish someone had told me some of them earlier.

Learn the skill the business needs next

When you’re self-taught, the list of things you could learn is endless. Branding, coding, copywriting, finance, hiring, ads, SEO, video. It’s easy to spend a year learning interesting things that don’t move the business.

The filter I use is simple: what’s the one thing stopping this business from growing right now, and what do I need to know to fix it? Learn that. Ignore the rest until it becomes the bottleneck.

When my stores needed customers, I learned paid ads. Not because ads were exciting, but because nothing else mattered until people showed up. Later, when the work needed to run without me, I learned to build software. Each skill arrived because the business asked for it.

Learn on real work, with real stakes

Courses and videos are useful for vocabulary. They tell you what things are called and what’s possible. But you don’t really learn a skill until you use it on something where the result matters.

So find the smallest real version of the work. Want to learn ads? Run a small campaign for something real, with a budget you can afford to lose, and watch every number. Want to learn copywriting? Rewrite your own product page and see whether more people buy. Want to learn automation? Automate one boring job you actually do every week.

Small, real and measured beats big and theoretical. You’ll learn more from one modest campaign you watched closely than from a long course about campaigns.

Keep the lessons cheap

The downside of learning by doing is that mistakes cost money. A few habits keep the cost down:

• Set a loss limit before you start. Decide what you’re willing to spend learning something, and stop when you hit it. Then work out what you learned before spending more.
• Change one thing at a time. If you change five things and it works, you’ve learned nothing you can repeat.
• Write down what you expected. Before you launch anything, write one line on what you think will happen. Afterwards, compare. That gap is where most of the learning is.
• Ask people who’ve done it, specifically. Not “how do I grow my business?” but “how did you decide when to stop a campaign?” Specific questions get useful answers, and people are more generous with their time than you’d expect.

Keep a record of what you learn

Self-taught knowledge is easy to lose, because nobody hands you notes. You learn something the hard way, move on to the next problem, and a year later make the same mistake in a different form.

A plain running document fixes most of that. Each entry needs only three lines: what I tried, what happened, what I’d do next time. It doesn’t need to be tidy. It needs to exist, and you need to read it before starting something similar.

Over time that document becomes your own playbook. It’s also the start of every process you’ll eventually hand to someone else or automate, so the effort pays twice.

Know when to stop teaching yourself

Learning everything yourself is a strength early on. It becomes a weakness when it slows the business down.

A rough test: if a skill is central to how your business wins, learn it deeply, because you need to judge other people’s work in it. If it’s necessary but not central, learn enough to know what good looks like, then pay someone or automate it. Your bookkeeping probably falls in the second group. Your offer and how you get customers probably fall in the first.

Paying for help isn’t giving up on being self-taught. It’s using what you taught yourself to choose the right help and check that it’s working.

Be wary of borrowed confidence

When you’re teaching yourself, it’s tempting to copy whoever seems most certain. There’s a whole industry built on certainty: screenshots, rented cars, guaranteed results.

Real operators tend to sound less certain, because they’ve seen what goes wrong. They talk about tests, not guarantees. They’ll tell you what didn’t work. If someone promises you a result before they’ve looked at your business, that tells you something about them, not about your business.

That’s one reason I keep my own claims plain and in my own words on this site. I’d rather you ask me about them on a call than take them on faith.

Build things that outlive your attention

The last lesson took me the longest. A self-taught founder tends to do everything personally, because that’s how they learned. That works until the business grows past what one person can hold in their head.

At some point the job changes from doing the work to designing how the work gets done: writing down processes, automating the repetitive parts, and handing over the rest. That shift is hard for people who taught themselves everything, because doing it yourself is what got you here.

It’s also where a lot of the growth is. My path went from stores to ads to apps and automations for exactly that reason.

If you’re starting out

You don’t need permission to learn this stuff. You need a real problem, a small budget for mistakes, and the patience to look at what actually happened instead of what you hoped would happen.

I talk about this with students, young founders and people changing careers. If you’re organising an event for that kind of audience, I’d be glad to speak. And if you’re already running a business and stuck on what to learn or fix next, book a call.