Building a website for the AI era

AI has become one of those variables that is difficult to ignore in business.

It is changing how we write, how we design, how we code, how people search for information, and how businesses automate parts of their work. Naturally, this also raises a lot of questions about websites.

• Do we still need them?
• Will AI build them for us?
• Does SEO still matter? 

There is a lot to unpack here, and I don't think there is one simple answer.

Partly because not every website has the same purpose. And partly because AI is changing some parts of the website ecosystem much faster than others.

So instead of trying to predict what websites will look like in five or ten years, I want to look at what is actually happening now.

Let's start with the people who have been building websites and helping businesses get found online for years.

What is going on with the experts?

Copywriters: from producing words to bringing something to say

Copywriting was probably one of the first areas where the impact of generative AI became obvious. Today, AI can write almost any kind of text, in almost any length or tone, almost immediately. A homepage, a sales page, an email, a blog post, 5 headline variations — you name it. And it's genuinely impressive. We shouldn't pretend otherwise.

But most of what it produces is still generic. It draws from the same enormous pool of information that everyone else has access to, so it can give you a good version of what is already known without giving you something that is specific to your business.

So does that mean we should stop using it? No, of course not.

We've gradually started separating 2 things. 

There are plenty of situations where we don't need a copywriter skill anymore. If you need a first draft or have a few ideas that need turning into coherent sentences, AI is an incredibly useful tool. It can make writing much more accessible to people who previously might have needed help with even the simplest pieces of copy. And it has become a genuinely useful tool for copywriters themselves.

But this doesn't mean the expertise of a copywriter disappears, because producing words has become easy. Having something worth saying hasn't! And this is where the copywriter's role shines.

A good copywriter doesn't just turn thoughts into sentences. They help uncover the thoughts worth turning into sentences in the first place. They know how to ask the questions that get beyond the generic description of a business, notice the interesting idea hidden in an ordinary answer, challenge vague positioning, and turn someone's experience into something another person can recognize and care about.

AI can help articulate those ideas once they're there. But it doesn't replace the process of discovering and shaping them.

Designers: AI can make something good. Can it make it yours?

Design has gone through a similar shift. AI can design. And it can design surprisingly well.

If you ask it for a clean, modern website for a consultant, a portfolio for a photographer, or a landing page for a SaaS product, you can get something that looks perfectly professional. For a lot of businesses, especially those that don't need a highly distinctive visual identity, this may already be enough.

And I think that's worth acknowledging: AI is making decent design more accessible.

But there's a difference between creating something that looks good and creating something that feels unmistakably connected to a particular business.

AI is very good at patterns. It has seen enormous amounts of existing design work, which means it knows the visual conventions that are currently considered attractive, modern, trustworthy, or professional.

That is also its limitation.

If you ask for a "beautiful website for a business coach," you're likely to get a beautiful version of something that already exists.

You can prompt it to be more unique, of course. You can spend an enormous amount of time refining prompts, uploading references, adjusting directions, and going back and forth. 

But originality isn't simply a matter of giving the right instruction.

A designer isn't only choosing between visual options. They are making decisions about what should feel familiar and what should feel unexpected, what reflects the personality of the business and what doesn't, which convention is useful and which one is worth breaking. Some of those decisions are difficult to explain even when you're making them yourself.

That's why I don't think AI is making designers irrelevant. I think it is changing what we need designers for. The deeper work of turning a business into a visual language is still a very human process.

And there is another layer to this that is easy to miss: the website isn't only the design you see.

AI can create something that looks polished while producing a messy build underneath. The more custom the design becomes — overlapping elements, unusual layouts, complex interactions, non-standard components — the more difficult it can be for AI to translate that idea into a clean and maintainable website.

That's important because the hidden layer affects things visitors may never consciously notice: performance, accessibility, responsiveness, SEO, and how easy the website will be to maintain 6 months from now.

SEO and AEO: from being found to being understood

Search is where things become particularly interesting, because AI is not only changing how websites are built. It is also changing how people discover them.

For years, SEO has largely been about helping search engines understand that a particular page is relevant to a particular query. We learned to think about keywords, headings, links, metadata, and all the other pieces that help a search engine crawl and interpret a website.

Those fundamentals haven't suddenly disappeared:.

• A fast website is still a good thing.
• An accessible website is still a good thing.
• Semantic HTML is still a good thing.
• Useful, original content is still a good thing.

But there is now another layer to consider.

People aren't only searching by typing a few words into Google. They are increasingly asking AI systems questions in a much more conversational way.

Instead of searching for:

"Webflow designer coaches"

someone might ask:

"Who could help me turn my outdated website into something that actually supports my coaching business?"

The second question contains much more context.

This is where the conversation around AEO — answer engine optimization — comes in.

We're still helping machines discover and understand our content. But the machines are becoming more sophisticated, and the way people interact with them is changing.

And it brings us naturally to the website itself. Because if AI is trying to understand your business, where is it going to learn about you?

Ideally, from you.

Your website can become one of the clearest and most complete sources of information about your business: what you do, who you help, what you believe, what you know, what you've done, and how you work.

And what about AI agents?

There's one more part of the AI conversation that inevitably reaches websites: what happens when AI starts doing things for us?

We've already had automations for years.

• A person fills out a form, and their information goes into the CRM.
• Someone books a call, and they receive an email.
• A payment is made, and an invoice is created.

Those systems are incredibly useful because they follow predictable rules. If X happens, do Y.

AI agents are different.

They can interpret information and make decisions. In theory, an agent could look at a new lead, decide what kind of person they are, determine what they need, update the CRM, write a personalized response, schedule a follow-up, and continue the process without a human touching it.

The possibilities are exciting.

But I think this is one area where a little caution is healthy. The problem is that agents can make mistakes in ways that traditional automations don't.

When an automation makes a mistake, you can usually trace it back to a rule.

When an AI agent makes a decision, the situation can be much less predictable.

For a large company with resources to monitor, test, and correct those systems, perhaps, the calculation may be different. But for a small business, where one incorrect email to a client or one incorrect CRM action can create a surprisingly large mess, the risk can easily outweigh the benefit.

So for now, I see a lot of value in a middle ground.

• Use AI to help you build and improve automations.
• Use AI to assist people in making decisions.
• Use AI for tasks where an imperfect first draft is perfectly acceptable.

But don't assume that because something can be automated, it should be handed over completely to an autonomous agent.

What is going on with websites?

Some websites simply won't change

Let's return to where we started.

Not every website needs to become a sophisticated AI-ready business system. Some businesses genuinely need a simple online presence, and that's perfectly valid.

You still need to explain who you are, what you offer, and how people can contact you. You still want it to look professional, work well on mobile, load reasonably quickly, and be accessible.

In fact, AI may simply make this type of website easier and faster to create.

If a simple website does the job, you don't need to turn it into an elaborate ecosystem just because AI exists.

The interesting changes happen with the other category of websites. The ones that are not simply there to prove that the business exists, but are actively involved in how the business works.

The role of other websites is evolving

For businesses that use their website as more than a digital business card, I think the change is much more significant.

I've always liked the idea of a website as the home of a business online. Not because it has to be the most important marketing channel, and not because every business needs to send everyone there, but because it is the one place that you actually own.

Your Instagram account isn't yours in the same way. Neither is your LinkedIn profile, your Youtube account, or your presence on any other platform. You are using someone else's infrastructure, following someone else's rules, and accepting the possibility that the platform, its algorithms, or its business model may change.

Your website is different.

You decide what lives there. You decide how it is organized. You decide what you want people to know about you. And, importantly, the knowledge you put there can accumulate over time.

I think AI makes this ownership even more valuable.

AI systems are increasingly trying to understand businesses, people, and topics in a much richer way. They aren't simply matching a keyword on a page. They are trying to form a picture.

Your website is becoming a knowledge base. Not in the technical sense of a documentation library, but as a place where the accumulated knowledge of your business lives: what you know, what you believe, how you work, what you've learned from clients, and how you see your field. Your articles, case studies, FAQs, resources, and service pages all contribute to this picture.

This also changes the way we think about content. We don't need to stop thinking about SEO or keywords, but there is more to build than a collection of pages answering popular search queries. If someone — or an AI system — looks at your website, you want them to get a sense of the whole picture. What topics do you talk about? What are your opinions? What connects your different ideas? What makes your approach different? In that sense, your website becomes less like a box filled with optimized articles and more like a world you've built around your expertise.

And because your business keeps changing, that world needs to stay alive. A website doesn't have to be constantly redesigned, but it should eventually evolve with the business. 

So, what actually changes in the AI era?

Some parts of creating websites are becoming easier to automate, and some can now be done without an expert at all.

At the same time, in professional work, AI hasn't replaced experts. It has become a tool that helps them work more efficiently, speed up repetitive tasks, and explore more possibilities.

And the more human part of their work still remains — and in many ways becomes even more valuable: human perspective, human thinking, creativity, experience, taste, and the ability to understand other people.

The same applies to search and automation. And while AI can make automations easier and more powerful, I still see traditional, predictable automations as more reliable than fully autonomous AI agents for many small businesses.

All of these changes affect how websites are created and maintained. But I think the biggest shift is happening in what the website itself is becoming. 

SEO isn't disappearing, but it is expanding as AI becomes another way people discover information. 

I wouldn't try to make my website "AI-proof." I'm not sure such a thing exists, and the technology will keep changing anyway. Instead, I'd invest in things that become more useful regardless of exactly how AI develops.

• I'd make the website clearer.
• I'd document what I know.
• I'd add real examples and case studies.
• I'd develop and share original ideas.
• I'd keep the information current.
• I'd make sure the technical foundation is solid.
• I'd connect the website to the systems that actually help the business run.
• And I'd use AI wherever it genuinely makes those things easier — without assuming that everything needs to become autonomous just because it can.

When anyone can produce a website in an afternoon, the question becomes less "Can I have a website?" and more "What do I want my website to say about my business?"

I don't think the answer to that question has changed because of AI. If anything, AI has made it harder to avoid answering it.