AI Can Build a Website in 60 Seconds. So What Am I Actually Selling?

I have been thinking about this question a lot recently, and if you are a web developer, freelancer or small agency owner, I think it is becoming increasingly difficult to ignore.

Open LinkedIn, Facebook, Instagram or YouTube and sooner or later you will see another advert promising something along the lines of:

“Build your website with AI in 60 seconds.”

Sometimes it is £10. Sometimes it is free. Sometimes it promises to create the website, write the copy, generate the images and optimise everything for you before you have even finished your coffee.

The uncomfortable part is this:

Some of these tools are actually getting pretty good but not perfect.

Not something I would blindly put into production and hand over to a client without checking it. However, good enough that pretending AI cannot build websites is no longer a serious argument.

As a developer, that forced me to ask myself a question I did not particularly like:

If AI can build a website in minutes, what exactly am I asking somebody to pay me for?

The Way I Used to Think About Web Development

Until recently, my business model for Client Flow Now was fairly straightforward.

1. Find a business that needs a website. (Maybe a plumber, an electrician, a cleaning company or a local tradesman).

2. Show them what I can build.

3. Explain why having a professional website matters.

4. Give them a price.

5. Hopefully win the project.

6. Build the website.

7. Get paid.

8. Repeat.

There is nothing fundamentally wrong with that model. Thousands of freelancers and agencies operate this way.

Unfortunately, it wasn’t working.

I could build the websites. That was never really the difficult part.

The difficult part was convincing somebody who may already have a Facebook page, a cheap website, a Wix account or simply no interest in improving their online presence that spending £1,000, £2,000 or more on a professionally built website was a worthwhile investment.

And now AI has added another layer to that problem.

The same business owner is being told that instead of paying a developer thousands of pounds, they can type a few sentences into an AI tool and have a website almost immediately.

I Started Wondering Whether I am still needed?

My first reaction to AI was probably similar to that of many developers:

• If something used to take me 40 hours and AI allows me to do it in 15, surely the value of that development has gone down.

• If AI can generate a hero section in seconds, why would somebody pay me to spend an hour building one?

• If it can generate a responsive layout, write JavaScript, create database models or help debug an application, what happens to the value of the developer?

Maybe the Product Was Never the Code

Recently, I completed a course by Swedish developer Nils Nilsson about building long-term relationships with creative agencies rather than constantly trying to sell individual websites directly to end clients.

I originally bought it because I was looking for another way of finding web development work. Instead, it made me reconsider what I am selling.

A client does not really care how many lines of code I write.

They do not care whether a particular component took me six hours or sixty minutes. They certainly do not wake up in the morning excited about my choice of JavaScript architecture. They care about the final result.

• Does the website work?
• Does it look right?
• Does it work properly on mobile?
• Is it fast?
• Can customers contact them?
• Can their team update it?
• Does it integrate with the systems they already use?
• Will it be delivered when promised?
• And when something goes wrong, can somebody fix it?

That is a very different product from:

“I sell web development by the hour.”

The Agency Perspective Changed Things for Me

This became even clearer when I started thinking about creative agencies. Imagine a branding or creative agency. They may be excellent at brand strategy, graphic design, content, marketing and dealing with their clients. They might also have access to exactly the same AI tools that I have. Technically, they could probably generate a website themselves.

But that does not necessarily mean they want to.

Someone still needs to understand the requirements, make technical decisions and turn the idea into something reliable. The same someone needs to deal with responsiveness, CMS integration, forms, APIs, performance, accessibility, deployment, hosting, bugs and all the other things that appear between:

Perhaps the value is simply being the person who can say:

“Leave the technical side with me. I will get it delivered.”

AI Might Actually Increase My Value

Imagine a carpenter before power tools.

If driving hundreds of nails into timber took hours with a hammer, the invention of the nail gun dramatically reduced the amount of time required. Nail gun did not suddenly make the hammer worthless. It made the carpenter more productive. I think AI could do something similar for developers.

The mistake would be measuring our value entirely by the amount of time required to complete the work. Imagine a project that previously required 40 hours at £75 per hour:

40 hours × £75 = £3,000

Now imagine AI-assisted development allows the same developer to deliver first results in 15 hours. If the only thing being sold is time, something strange happens:

15 hours × £75 = £1,125

The developer became dramatically more productive and their reward was a 62.5% reduction in revenue. That does not make much sense. Business receiving the finished project may still receive exactly the same value.Possibly more, because the project was completed faster.

The thing that changed was the developer’s capacity.

Capacity May Be the Real AI Opportunity

The conversation around AI and development is usually framed as:

Atificial Intelligence vs. Developers.

I am beginning to think a much more useful question is:

What can one good developer now accomplish with AI that previously required two or three people?

If a developer previously had enough capacity to deliver two projects in a month, perhaps AI-assisted development allows them to deliver three or four.

That does not necessarily mean firing developers or replacing an entire team with AI. It means the point at which a business needs to hire additional people could move significantly.

For a small development business, that could be huge.

Instead of immediately solving increased demand by adding another developer, you can first improve your systems, automate repetitive work and use AI to increase your existing capacity. Only when that capacity is genuinely exhausted do you need another person and when you eventually do hire somebody, they can use the same tools.

There Is a Catch

None of this works if AI is used as an excuse to produce rubbish faster - generating code is not the same as delivering a production-ready project.

AI can confidently produce code that looks completely reasonable while introducing poor architecture, accessibility problems, security issues, unnecessary dependencies or bugs that only appear under particular conditions.

The developer’s role therefore starts moving.

• Less time may be spent writing every line manually.
• More responsibility moves towards:

understanding, directing, reviewing, testing and making decisions.

The skill is no longer simply:

“Can you write this code?”

It becomes:

“Can you use everything available to you to deliver this properly?”

That includes AI.

This Also Changes How I Think About Pricing

I am not suddenly claiming that hourly pricing is dead. There are plenty of projects where charging by time makes sense.

Howeverm AI exposes a weakness in treating hours as the only measurement of value.

If experience, reusable components, better processes and AI allow one developer to solve a problem faster than another, charging purely by the hour can effectively punish efficiency. That has made me think much more seriously about project pricing and value.

What is the client actually buying? Twenty hours?

Or a functioning website that solves a business problem?

Those are not necessarily the same thing - developers need to become better at explaining the second one.

So What Am I Actually Selling?

If AI can build a website in 60 seconds, what am I actually selling?

I do not think I have the complete answer yet but I know my answer is changing. I am not selling the number of hours I spend in VS Code, nor selling the number of lines of code I can write.

And increasingly, I do not think I should even be selling “a website” as the entire proposition.

What I can sell is the responsibility for taking a technical problem from an idea to a finished result. The experience to know when AI has produced something useful and when it has produced nonsense. Ability to make decisions when the project does not go exactly according to plan and communicate with the people involved.

Ultimately: confidence that somebody owns the technical side of the project and will get it delivered.

It might allow me to deliver more of it. That leads to the second question I am now trying to answer. If constantly searching for individual businesses that need a website is not the best model anymore…

What happens if I stop looking for website clients and start looking for businesses that already have them?

That is what I am going to test next.