Skip to content
Currently booked
· 7 min read

Why not just build it yourself with AI? A word about haircuts

AI tools can turn a prompt into a working project. So why hire someone? The same reason most people don't cut their own hair: what AI does well, what you can't see yourself, and when doing it yourself is fine.

Request an AI summary

AI tools can now turn a few sentences into something that runs. So it’s a fair question: why pay someone, when you could just ask the AI yourself?

My honest answer is a haircut.

1. The haircut you could do yourself

You could cut your own hair. Clippers cost less than a single haircut, there are countless tutorials, and the mirror is right there. Most people still go to a hairdresser/barber.

Not because cutting hair is magic, but because the result matters, and because of the back of your head. You can’t see it, you don’t know what good looks like from behind, and by the time you notice, it’s too late. In the best case, the result is just not quite what you had in mind. In the worst case, you’re wearing a hat for three weeks.

Building a project with AI works much the same way.

2. Not just haircuts

We hand off plenty of things we could do ourselves: for a better result, to save time, or simply to treat ourselves now and then.

  • Photos. Your smartphone takes great pictures. For the ones that really count, like a wedding or your team on your own website, most people still book a photographer.
  • Moving house. You can rent a van and carry every box yourself, or have movers do it while you get on with your week.
  • Taxes. You can fill in the forms yourself, or have an adviser who knows which deduction you forgot.
  • Building. You might put up a garden shed yourself. For a house, you bring in people who know about foundations and the wiring behind the walls, and who leave room to add a floor later.
  • The car. You can change the brake pads with a video next to you, but you want them to work on the motorway.

For a business, time is money in the most literal sense. Every hour you spend learning to build something is an hour you don’t spend with your customers. Often the question isn’t whether you could do it yourself, but whether that’s the best use of your time.

3. What AI does well

AI is a great tool, and I use it every day myself. If you want to try an idea, build a small tool for yourself, or see what’s possible, go for it. You’ll learn a lot and you’ll get something that runs.

The front of the haircut usually looks fine at first glance: the part you see straight away, on your screen, in the demo. Both the front and the back deserve a closer look, though.

4. The front: looking good isn’t the whole story

Even the part you can see deserves a second look. When you’ve built something yourself, you know where everything is and what each button does. Your visitors don’t. And whether something looks good is only half the question; the other half is whether it works for the people using it. That’s user experience, or UX, and most of it happens behind the visuals:

  • The path through it. Do people find the one thing they came for within seconds, or do they have to search? Is the main action obvious, or does it compete with five others?
  • The moments in between. What people see while something loads, when a list is still empty, or when they make a typo in a form. Good error messages and button labels decide whether someone finishes or gives up.
  • Thumbs, not mouse pointers. Buttons big enough to tap, forms that bring up the right keyboard, and nothing important hidden behind hover effects that don’t exist on a smartphone.
  • Everyone, not just you. Enough contrast to read in sunlight, text that can be enlarged, and a layout that still makes sense with a screen reader.

Taste can play tricks, too. Something can look great to you and still feel dated to your visitors, because it’s what was in fashion a few years ago:

  • The big slider on the homepage that cycles through five messages, when most people only ever see the first one.
  • An intro or splash screen people have to click away before they get anywhere.
  • Pop-ups on arrival asking for a newsletter sign-up before anyone has read a single line.
  • A hamburger menu on a wide desktop screen, hiding navigation that would easily fit.
  • Stock photos of handshakes and people pointing at laptops.

And there’s a newer one: the AI look. Ask an AI for a design and you often get the same thing everyone else gets: the same soft gradients, the same rows of rounded cards with an icon on top, the same headline patterns. It looks polished at first glance, but it makes your brand look like everyone else’s.

This is where I help: I look at the front with fresh eyes, test it the way your visitors will use it, and make it look like you, not like the template everyone else started from.

5. The back of your head: what I take care of

Some parts of a project don’t show in a demo, and they’re exactly the parts I cover for my clients, whether we start from scratch or from something you’ve already built:

  • Security. Logins, forms, uploads and payments set up to hold up, and kept that way.
  • Data and privacy. Customer data stored in the right place, reachable only by the right people, and in line with what the law expects.
  • Accessibility. Usable with a screen reader, with just a keyboard and with poor eyesight, so nobody is left out.
  • Performance. Fast on an older smartphone over a weak connection, not just on your own device.
  • Being found. Built so search engines and AI assistants understand what it is and who it’s for.
  • Growth. Planned so it still holds with ten times as many users, and new features fit in without starting over.
  • Maintenance. Updates, backups, and someone who knows how it all works when something needs to change.

You focus on your business and your customers. I make sure the back of the head looks as good as the front.

6. What you’re paying for

When you hire someone, you’re not paying for the typing. AI does a lot of that now, and I’m open with my clients about how I work with it. You’re paying for:

  • Knowing what to build, and what to leave out.
  • Judging the result: spotting the mistake that looks right.
  • Planning ahead, so the project scales and still holds up in a few years.
  • Taking responsibility for it, including the back of the head.

The tool is only as good as the person holding it. A hairdresser/barber with great scissors still needs to know how to cut.

7. When doing it yourself is fine

There’s no shame in cutting your own hair at home, and I always think it’s a good sign when a client wants to try it themselves first. Don’t let this post scare you off: you learn what’s possible, you find out what you actually want, and our conversations get better for it. Nothing you’ve built is wasted either; it’s often a great starting point.

Doing it yourself works well for:

  • Experiments and prototypes, to find out whether an idea works at all.
  • Personal projects without customer data or payments.
  • Getting a feel for what you want, before you talk to someone.

Once real customers, their data or your business depend on it, it’s worth having a professional take a look, at least at the back of the head.

The short version

  • AI can build a lot, and it’s a great tool for trying ideas.
  • Looking good isn’t the same as working well: UX, dated trends and the generic AI look are easy to miss in your own project.
  • I cover the parts you can’t see yourself: security, data, accessibility, performance, growth and maintenance.
  • You pay a professional for judgement, planning and responsibility, not for the typing.
  • Handing it over gets you a better result and saves you time, and for a business, time is money.
  • Do it yourself for experiments; bring someone in before it goes live with real customers.

If you’ve started something yourself and would like a second pair of eyes, or would rather hand it over from the start, get in touch.