Can AI Find You? What WordCamp US Taught Us About Building for the Agentic Web

WordCamp US 2026 was a WordPress community event, and the part that stayed with me was how many conversations came back to AI’s impact on website building and a simple question: if an AI agent comes looking for your business, have you built your site in a way it can actually understand?

I had the opportunity to attend it in Phoenix last week, and one session I’ve kept thinking about was “From Crawlers to Customers: WordPress and the Rise of AI Buyers,” presented by Danny Dudzic and Shani Banerjee.

The session was largely about agentic commerce and what happens when AI moves beyond helping someone search for a product and can actually begin interacting with a store on that person’s behalf.

While I was listening, I kept thinking about something that happens before the transaction ever becomes relevant.

Can an AI agent find you, and can it actually understand what it finds?

That question is bigger than ecommerce. It applies to service businesses, healthcare organizations, manufacturers, restaurants, local companies, and really anyone relying on a website as part of how people discover and interact with their business. If you design, build, manage, or market websites, that’s no longer a theoretical problem. It’s becoming a visibility problem as AI-driven search changes how customers find and evaluate businesses online.

We’ve spent years thinking about whether people can find a website. Then whether search engines can crawl and understand it. More recently, we’ve started thinking about how businesses show up in generative search.

AI agents add another layer because they aren’t necessarily searching the way we do.

Someone doesn’t have to figure out the perfect combination of keywords first. They can just describe what they need, which changes what the website has to communicate.

This post looks at what WordCamp US 2026 revealed about that shift, how AI search differs from traditional search, why so many of the decisions that shape AI readability happen during the initial build, where accessibility overlaps with machine understanding, where friction shows up in AI-driven commerce, and what building websites for the agentic web actually looks like in practice.

Search has more context now

Think about the difference between searching for a knee brace and asking an AI to find a knee brace that’s appropriate for running and provides lateral support.

Or searching for a healthy breakfast versus asking for something high in protein, dairy-free, and low in added sugar.

Neither request is particularly complicated. There is just more context built into them.

With the knee brace, the system needs enough information to understand what makes a product appropriate for running and whether it provides the type of support being requested.

With the breakfast search, “healthy” isn’t really useful enough. The system needs ingredients and nutritional information to determine whether something actually meets the request.

The same thing happens with services. Someone might ask for a provider who performs a particular service in their area instead of searching the exact phrase we’ve optimized a page around.

AI search tools and agents have to understand the business well enough to make that connection on their own. That means we are moving beyond whether the right words happen to exist somewhere on the page. The site needs enough context to explain what the business does, who or what a service is for, where it’s available, and how it relates to what someone is actually looking for.

And that is where I started connecting this back to the way we build websites in the first place.

Trew Knowledge’s WordCamp US 2026 recap captured this well: the idea of the site as a source of truth grows more important when content may appear in search results, AI answers, conversational interfaces and agent workflows. Content cannot depend entirely on visual presentation to communicate meaning. Relationships, metadata, reusable components and consistent information architecture become part of how an organization’s knowledge is understood outside individual pages.

A lot of this is decided before SEO ever touches the site

SEO and GEO are specialized disciplines. I don’t think web designers need to suddenly take those jobs over, nor do I think every website project needs to solve an entire future growth strategy before it launches.

Build strategists make a surprising number of decisions during the initial build that determine what those specialists will have to work with later.

We decide the information architecture and how services are organized. We create page relationships and CMS structures. We decide what becomes a field, what that field is called, whether something deserves its own page, and how information is grouped together.

We determine the heading hierarchy, build navigation and conversion paths, and decide whether important information exists as actual content or whether its meaning is being carried almost entirely by the visual design.

Those choices aren’t necessarily SEO work, but they absolutely affect SEO and GEO work later.

If the structure is good, another specialist can come in and build on it.

If it isn’t, growth starts with cleanup.

Services that need room to stand on their own can end up buried inside a single page. CMS relationships may not clearly connect locations, services, or other related content. Business information can become inconsistent across the site, and adding something new may mean manually rebuilding the same structure again and again.

None of those problems are impossible to fix, but now the growth team has to spend time fixing the website before they can actually work on growing it.

That is avoidable.

A traditional web build can still create a strong foundation by getting the basics right from the start: clear information architecture, semantic structure, accessible content, clean URLs, thoughtful CMS relationships, all the things that make the site easier to build on later.

That doesn’t mean we solved SEO or GEO during the build.

It just means we didn’t make their jobs harder.

Clarity is doing more work than it used to

One thing I liked about a recurring theme at WordCamp was how quickly conversations about AI came back to fairly ordinary things. A lot of it was really about simplicity: accurate data, good structure, useful metadata, consistent naming, and fewer places for meaning to get lost.

Not exactly the flashy side of AI, or even a web build.

But it makes sense. People are good at filling in gaps. We can look at a photo, skim a page, and piece things together when the information is messy. AI needs us to be more explicit.

For ecommerce, that might mean clear product details, sizing, ingredients, or inventory. For service businesses, it could mean clearly defined services, locations, or credentials. The specifics change by industry, but the relationships between things still need to make sense.

That’s also why GEO feels like a natural extension of this conversation. The useful information is around the keyword: what the thing is, who it’s for, what attributes matter, when it is appropriate, and how it differs from another option.

Keywords still matter. They just aren’t carrying the whole meaning by themselves.

And I think that is probably closer to what good web content should have been doing all along.

Accessibility fits into this conversation for a similar reason.

An AI agent isn’t navigating a website by looking at it the way a sighted human does. It relies much more heavily on the meaning and structure underneath the visual presentation.

Accessible web design has been telling us for years that information shouldn’t only make sense because of how it looks.

Headings need meaning. Labels need meaning. Links need to make sense in context. Relationships between information can’t depend entirely on color, position, or visual hierarchy.

But it is interesting to see the same principle surface from another direction. When the meaning of a page exists beyond its visual presentation, more people and more systems can understand it.

Miriam Schwab’s WordCamp US session “The Agentic Web Is Coming. Is WordPress Ready?” pointed directly at this: AI agents don’t browse the web the way humans do. They need to read it, query it, take actions on it. WordPress, which already powers 43% of the web, has a real structural role to play here. But it requires the ecosystem to think differently about what a WordPress site actually is.

Friction doesn’t disappear. It moves.

The commerce side of the talk pushed all of this one step further.

If an AI agent can eventually act on someone’s behalf, then understanding the website is only part of the job. It also has to understand what can happen next.

That might mean making a purchase, booking a service, or simply knowing what the next step is.

We already spend a lot of time in web design reducing friction for people. We simplify navigation, forms, checkout processes, and conversion paths.

AI agents will have their own version of friction.

The problem may not be that a button is difficult to find. It could be that the data is unclear, two pieces of information conflict, a relationship hasn’t been defined, or there isn’t an obvious way for another system to complete the next action.

In an AI-driven search experience, that kind of ambiguity can be enough to keep a business from being the answer.

As The Repository noted in their WordCamp US 2026 coverage, at least 18 sessions across the schedule were AI-focused or AI-adjacent, making it the largest AI showing at a flagship WordCamp to date. The sessions covered agentic development, AI-powered plugin pipelines and whether WordPress is ready for the agentic web. That breadth signals something: this is not a niche developer concern. It is becoming a core question for anyone building or maintaining a website.

Building for what comes next

I don’t think anyone knows exactly how traditional search, generative search, AI agents, and websites will work together a few years from now. We probably don’t need to.

What we can do is build websites that aren’t fighting whatever comes next. Clear structure, accessible content, consistent data, enough context to make the information useful, the basics still matter.

That gives a growth strategist something solid to build on instead of something they have to untangle first.