AEO vs. SEO in 2026: What Colorado Businesses Actually Need to Do

If your website traffic is down this year but your phone is ringing about the same, you are not imagining it and you are not broken. Google changed how it answers questions, and the number that used to measure your visibility stopped measuring it. Here is what actually changed, and what is worth doing about it.

Every week lately we get a version of the same call. A business owner opens their analytics, sees organic traffic down twenty or thirty percent from last year, and wants to know what went wrong. Usually nothing went wrong. Their rankings are intact. Their leads are flat or slightly up. What changed is that Google started answering the question on the results page instead of sending the visitor to the page that had the answer.

That shift has a name now, or several: answer engine optimization, generative engine optimization, AI search optimization. New vendors are selling it as a separate service with a separate invoice. Before you buy one, it is worth understanding what the data actually shows, because the honest version of this story is less dramatic and more useful than the version in your inbox.

What Actually Changed in 2026

Two things happened this year that matter to a local business. In May, Google made AI Mode the default search experience worldwide, running on a faster Gemini model and built around a conversational search box rather than a list of ten blue links. And the share of searches that end without anyone clicking anything kept climbing.

The first two numbers are why your traffic chart looks the way it does. Zero-click searches were around 60% in 2024 and reached 68% in the first four months of this year. When an AI Overview shows up, clicks to the results below it fall by roughly 60%.

The third number is the one almost nobody leads with, and it is the most important one for your budget. When researchers looked at nearly two million AI Overview citations, about 76% of the cited pages were already ranking in Google's top ten for that query. AI is not discovering a secret new set of websites. It is mostly reading the same pages that were already winning, and summarizing them.

Why Colorado Service Businesses Feel This First

AI answers are not distributed evenly across industries. A study of more than 60,000 local business queries found AI Overviews appearing in about 40% of them, but service categories sat at the top of the range. Cleaning services, carpet cleaning, and window cleaning all triggered an AI answer on roughly two out of three searches. Navigational searches, where someone is looking up a specific business by name, almost never did.

Two patterns in that data are worth knowing because they describe exactly how your customers search:

That split is the practical takeaway. The top of your funnel, where people are researching and comparing, is where you are losing clicks to AI summaries. The bottom of your funnel, where someone is ready to call a plumber in Parker tonight, still looks much like it always did. If your traffic is down but your lead volume held, this is almost certainly why.

What Actually Earns AI Citations

In May, Google published its own guidance on performing well in AI Overviews and AI Mode. The headline from it is blunt: there are no additional requirements to appear in these features, and you do not need to create new machine-readable files, AI text files, or special markup. The recommendations are the fundamentals — let crawlers through, keep important content as text, link internally, keep structured data consistent with what is visible on the page, and keep your Business Profile current.

That is not a reason to do nothing. It is a reason to do the right things. Here is where we focus for the clients we manage:

• Answer the question in the first three sentences. AI systems lift the passage that most directly answers the query. A page that spends four paragraphs warming up before getting to the cost, the timeline, or the answer gets skipped in favor of one that leads with it.
• Publish the specifics your competitors hide. Real price ranges, real timelines, real conditions where a repair beats a replacement. Pages with concrete, checkable detail get quoted. Pages full of “quality workmanship since 2009” do not.
• Build one strong page per service per city, not one page for everything. The old local SEO rule survives intact, and it matters more now because the AI needs to match a specific question to a specific page.
• Treat your Google Business Profile as primary infrastructure. Hours, categories, service areas, and photos feed local AI answers directly, and Google names it explicitly in its own guidance. Review volume and recency do real work here.
• Get mentioned where the AI already reads. Reputable local directories, chamber and association listings, local news coverage, and review platforms. AI answers are assembled from multiple sources, and third-party corroboration is part of what makes a business look real.
• Check that your crawlers are not blocked. We regularly find sites where a CDN, security plugin, or a stray robots.txt rule is blocking the crawlers that feed AI answers. This is the single most common fixable problem we see, and it is usually a ten-minute fix.
• Write FAQ sections in the words customers actually use. Not “What is our service area?” but “Do you come out to Elizabeth?” Conversational questions match conversational queries.

What We Would Not Pay For

This is where a lot of marketing budget is about to get wasted, so we want to be specific.

• 📄

  An llms.txt file for Google visibility

  Google's own documentation states plainly that llms.txt is not needed for AI Overviews, AI Mode, or any other generative search feature, and John Mueller has compared it to the old keywords meta tag. It has some legitimate uses for developer documentation and AI agents, but if someone is selling it to you as the key to showing up in Google's AI answers, that is not what the evidence supports.

• 🏷️

  “AI-specific” schema markup

  There is no special structured data for AI features. Standard schema that accurately matches your visible page content is useful and always was. A vendor charging extra for AI-flavored markup is selling you something that does not exist.

• ✂️

  A full-site rewrite “for AI”

  Chunking your content into AI-friendly fragments is on Google's list of things that do not help. Rewriting a page because it buries the answer is worth doing. Rewriting your whole site because a tool gave you a low AI readiness score is not.

• 📉

  An AI visibility score with no connection to leads

  Plenty of new tools will sell you a dashboard number. Some of them are genuinely useful for tracking whether you get mentioned. None of them are useful if nobody has connected that number to how many people actually called you. Track the citations if you want, but judge the program on booked work.

How to Check Where You Stand in 20 Minutes

You do not need a tool to get a first read on this. You need a notepad and twenty minutes.

1. Write down the ten questions you get asked most on the phone. Not keywords. Actual questions, in the words customers use. “How long does a roof take?” “Will my insurance cover this?” These are the queries most likely to produce an AI answer about your category.
2. Search each one, and add your city to half of them. Note whether you get an AI answer or a traditional list. Note which businesses and websites get named in the AI answers. That list is your real competitive set now.
3. Ask ChatGPT and Google's AI Mode the same questions. “Who should I call for a furnace repair in Castle Rock?” See whether you come up, and see what the answer says about you. Businesses are often surprised by what the AI believes about their hours, services, or service area.
4. For every question where you were not mentioned, check if you have a page that answers it. Usually the answer is no, or the answer exists buried on page four of a service page. That gap list is your content plan for the next quarter, in priority order.
5. Check your analytics for AI referral traffic. In GA4, look at your traffic sources for ChatGPT, Perplexity, Copilot, and Gemini. The volume is usually small. Look at what those visitors do once they arrive, because that is the number that tells you whether this traffic is worth chasing in your market.

Change What You Measure Before You Change What You Spend

The hardest part of this shift is not technical. It is that sessions stopped being a good proxy for visibility. If 68% of searches end without a click, traffic can fall while your actual presence in the market holds steady or grows. We have seen accounts where organic sessions dropped year over year and booked jobs went up, and the owner was convinced their SEO was failing right up until we put the lead numbers next to the traffic numbers.

So before you buy a new service, change your scoreboard. Track calls and form fills by source. Track branded search volume, because people who read about you in an AI answer often look you up by name afterward, and that shows up as direct or branded traffic rather than as a click from the AI. Track how many of the questions you listed above now have a page that answers them well. Those three things tell you far more about your position than a traffic line ever did.

The Bottom Line for a Colorado Business

AI search is a real change and the traffic impact is real. But the response that works is not a new line item. It is doing local SEO properly, writing pages that answer real questions directly and specifically, keeping your Business Profile and reviews in good shape, making sure nothing is blocking the crawlers, and measuring leads instead of sessions.

If you are already doing those things well, you are most of the way to being the business the AI names. If you are not, no amount of AI-specific tooling will get you there, because the engines are overwhelmingly quoting the pages that already earned their position the ordinary way.