Skate to where the puck is going

"I skate to where the puck is going to be, not where it has been." — Wayne Gretzky

As kids who played hockey, me and my teammates had one hockey hero: Wayne Gretzky. The man understood something fundamental: the best players don't chase. They anticipate.

As an adult that's been building websites for 29 years, I've seen websites, search engines and technology shift from one end of the ice to the other. From the animated gif to Flash to HTML5. Search engines from Alta Vista to Google. Desktop to mobile. And now we're experiencing the biggest shift yet as AI affects not only how websites are used, but also how websites are found.

This is not the time to be stuck at the wrong end of the ice.

Two shifts are happening at once. First, AI agents are becoming a new visitor persona. They crawl your site, parse your content, and decide whether to cite your content as a source. Attracting them requires a new approach: AIO, or Artificial Intelligence Optimization. AI agents and traditional search engines and bots do now work the same way.

Your SEO of yesterday does not cover your AIO for today... or tomorrow.

Second, the human experience is changing too. AI chatbots will soon become the primary interface, out performing navigation menus and search bars for engaging with websites. The businesses paying attention to where the puck is going, not where it's been, will be the ones in the right position to score.

Two forces are converging

Force #1: AI is changing how people use websites

Your visitors want answers, not browsing sessions.

Think about how inefficient we've been using websites. We hunt and peck through complicated mega navigation menus or use search tools that have no intelligence. That's changing. The AI chatbot is a massive shift that has yet to even be realized. It's still a bolt-on (literally). What you will soon see is that brands that want to put pucks into the net will be placing that chatbot right in front of the goal.

That's where your customers are heading. The interface is shifting from navigation menus to conversation. Chat is the starting point, but it's evolving fast: more visual responses, voice interaction, persistent "assistant" areas on sites, guided decision flows that feel less like search and more like a concierge.

This is happening now on millions of devices. Your future customers are already getting comfortable with it.

Force #2: AI is changing who the visitor is and how they find your site

Traditional organic search traffic is under real pressure.

AI summaries appear above search results. "Answer-first" experiences give people what they need without clicking through. Zero-click behaviors mean someone gets their question answered and never visits your site at all. And increasingly, the "visitor" isn't even a person. It's an AI agent retrieving information on their behalf.

The numbers are sobering. Rough estimates show most brands are losing organic traffic. The lucky ones are down only 15%. Many are losing over 50%.

This changes the game. You're not just competing to rank. You're competing to be the most cite-worthy, most machine-readable, most authoritative source. AI systems need to trust your content enough to surface it as the answer.

If your site is a tangle of inconsistent information, poor markup, and outdated pages, AI won't trust it. Neither will customers.

The new front door is a conversation

Navigation and on-site search will matter less.

They won't disappear. But they'll stop being the primary interface for many visitors. Navigation becomes a fallback. Search becomes a utility, not the strategy.

The chatbot becomes the main front door.

More people will start with "Ask" instead of "Find." And a good chatbot won't be a glorified FAQ page. It will be a guide that can clarify intent, route visitors to the right path, surface structured answers, and support actual decisions: comparing options, checking availability, completing transactions.

We're in phase one of this evolution. Current chatbots mostly handle customer service queries trained on FAQs. But as these tools integrate with backend systems (CRM, inventory, booking, ecommerce) they become something different entirely. They become agents that can take action, not just answer questions.

Imagine a customer conversation like this:

"I'm looking for a product similar to what I bought last year, but in a different size. Do you have it in stock? What's my best shipping option to get it by Friday?"

A well-built chatbot, connected to your systems, handles that entire interaction. No page-hunting. No confusion. Just answers, options, and a clear path to purchase.

That's where this is going. The question is whether your website is built to support it.

The new foundation: structured content, not pretty pages

Content modeling beats page design.

The businesses that win will be the ones producing clean, structured answers to the questions customers actually ask: hours, pricing, policies, availability, locations, shipping, returns, specifications.

This requires a content model, not just pages. "Pretty pages" without structured data become less valuable when AI is trying to extract a clear answer.

Think of it this way: if your return policy is buried in a paragraph on page 47 of your site, and formatted inconsistently across three different pages, AI can't reliably cite you. Neither can a chatbot. Neither can your own customer service team trying to give accurate information. When the same answer lives in three different places with three different wordings, AI doesn't know which to trust.

AI rewards authority and punishes ambiguity.

AI systems prefer content that is explicit, consistent, timestamped where it matters, and clearly marked as official. Businesses need to act like publishers of record, not just marketers with a website.

Accessibility, SEO, and AI readiness are the same thing now

The easiest way to understand why WCAG accessibility matters for AI is to think about screen readers.

Screen readers help blind and visually impaired people use websites. They don't "see" the page like you do. They consume a structured version of your code called the Accessibility Tree. It strips away the visual noise and exposes what matters: headings, links, buttons, form fields, their labels, and their relationships.

AI agents work the same way. They consume the Accessibility Tree. OpenAI has publicly stated that Atlas "uses ARIA tags, the same labels and roles as screen readers, to interpret page structure and interactive elements."

Build a site that works for screen readers, and you've built a site that works for AI agents. Same structure. Same investment.

The data backs this up. A SEMrush study of 10,000 websites found that WCAG-compliant sites saw a 23% increase in organic traffic. In benchmark testing, AI agents using accessibility data completed tasks successfully about 85% of the time. Agents relying on visual interpretation alone? Around 50%.

Note: I took a deeper dive into the questions Do Accessible Websites Really Perform Better for AI Agents on our sister site, Accessibility.Works. 

 

The new SEO is AIO, and accessible WCAG sites fuel both

SEO isn't dead, but it's evolving. The priority is shifting from keyword optimization to machine comprehension.

Traditional SEO focused on helping search crawlers find and rank your pages. AIO focuses on helping AI agents understand, trust, and cite your content. The goal is different. You're not just trying to rank. You're trying to be the authoritative source that AI pulls from when answering questions.

What does AIO look like in practice? Semantic HTML that clearly signals content structure. Schema markup that gives AI explicit context. Proper heading hierarchy so agents can navigate to relevant sections. Labeled form fields so agents can complete tasks. Internal linking patterns that reinforce meaning.

This isn't a new discipline. It's what accessibility practitioners have been doing for decades. Build for screen readers, build for AI agents. Same tree. Same payoff.

Stuffing keywords into meta tags won't cut it anymore. Google's AI overviews and competing AI search tools are looking for content they can understand, trust, and cite. That requires structure, not tricks.

Skate to where the puck is going

AI is shifting the game for websites. User expectations are changing. Discovery is changing. The interface itself is changing.

The businesses that will be positioned for what's coming are the ones building structured, accessible, authoritative systems now. That means treating accessibility as foundational architecture, not an overlay. It means building content models around the questions customers actually ask, not just the pages you think they should browse. It means creating a single source of truth that feeds your website, your chatbot, your customer service team, and the AI agents that will interact with your content.

The ones still thinking in terms of "pages" and "navigation" and "we'll update it next year" are going to struggle when AI becomes the primary way customers find and evaluate their options.

You've got time. But not as much as you think.

Propeller Media Works has been building websites for 29 years. We specialize in nex-gen SEO/AIO smart accessible web development, and helping businesses prepare for what's next. If you want to talk about what AI-ready looks like for your site, reach out.