When I launched photographer.com in 1995, search looked nothing like it does today. AOL was how most people got online, and Yahoo’s directory was where they went to find things. Google did not exist yet; it would not arrive until 1998. And yet, if you searched for a Washington, DC wedding photographer, my site came up first. I was doing SEO before I knew the term, and before the industry had invented it.
Years later, working at dBurns Design in Santa Monica, I did it deliberately. We dominated the results for anything related to “web design in Los Angeles.” The tactics that got us there, stuffing keywords into meta tags and page copy, were state of the art at the time. Most of them would do nothing today, and a few would get you penalized. Meta keyword tags, once a genuine lever, have been ignored by Google for well over a decade.
So I have watched this discipline change for thirty years. Here is what that vantage point makes clear. The strategy has held. The implementation has been rebuilt several times over. What is driving the current rebuild is not a new algorithm tweak. It is artificial intelligence changing where answers get delivered, and who does the delivering.
Three eras, and how far the ground has moved
Fifteen years ago, around 2011, SEO was largely a game of inputs you could count. Keyword density on the page. The raw number of inbound links. Exact-match domain names. Teams wrote content for the algorithm first and the reader second, and it often worked. Google’s Panda and Penguin updates were just beginning to punish that approach, but the mental model was mechanical: pull the right levers, watch the rankings move.
Ten years ago, around 2016, the ground shifted toward intent. Google’s RankBrain started interpreting what people meant, not just what they typed. Mobile traffic overtook desktop, so page speed and mobile-friendliness became ranking factors rather than nice-to-haves. The link-counting mindset gave way to a harder question: does this page actually satisfy the person who searched?
Five years ago, around 2021, the story was about experience and credibility. Google’s page experience signals rewarded fast, stable, secure pages. Its guidance leaned hard on E-E-A-T, meaning experience, expertise, authoritativeness, and trustworthiness. Featured snippets and “position zero” changed the goal from ranking first to being the answer. A growing share of searches ended without a click, because the answer sat right on the results page.
Each of these eras made the last one’s tactics look naive, the same way meta tags in 1995 look naive now. But notice what carried through all of it.
What never actually changed
Strip away the tools and the same three fundamentals sit underneath every era, from my first site in 1995 to an AI answer in 2026.
Know your audience, then go deeper than a keyword. This is the one that matters most, and it is the one that has not moved an inch. Identify who you are actually trying to reach. Segment them, because a procurement officer and an end user are not the same buyer. Then understand each segment’s needs, their pain points, the job they are trying to get done, and the exact way they would search for a solution and for the vendors who provide it. In 1995 I called this knowing my customer. Today the tools call it audience and intent modeling. It is the same discipline, and everything else is downstream of getting it right.
Earn trust and authority. Thirty years ago that meant being the obvious best result. Fifteen years ago it meant links. Today it means being the source that credible people, publications, and now machines choose to cite. The currency changed. The principle did not.
Remove friction. A site that is fast, accessible, well structured, and easy to navigate has always won. That was true when the concern was dial-up speeds, and it is true now that the concern is a crawler deciding whether your content is clean enough to quote.
If a vendor tells you the fundamentals are dead, be skeptical. The fundamentals are the one thing that has survived every shift. What they are almost certainly selling you is a new implementation of the same three ideas.
What genuinely changed: the tools, the techniques, and AI
The implementation layer is where the real disruption lives, and this cycle is the sharpest one yet.
The tools grew up. Search consoles, crawl and audit platforms, and analytics suites now surface issues in minutes that used to take a specialist weeks. More recently, those same platforms embedded AI that drafts content, clusters topics, finds internal linking gaps, and predicts which pages are at risk. The work that once justified a large team is increasingly assisted or automated.
The techniques matured. Modern search rewards topical depth over single keywords, structured data that machines can read, and content that demonstrates real experience rather than repeating a phrase. The craft moved from optimizing a page to building authority across a subject. The meta-tag tricks that won me “web design in Los Angeles” have no equivalent shortcut today.
And then AI moved the finish line. Google’s AI Overviews and AI Mode now answer many queries directly on the results page. Tools like ChatGPT, Claude, Perplexity, and Gemini answer questions that people used to type into a search box, and they often do it without sending a single click to a website. This is the part that should have a CMO’s attention, because it changes the fundamental transaction. For years the goal was a click to your site. Increasingly, the goal is a mention inside an answer that never links to you at all.
AEO and GEO: the new front, not a new religion
Two acronyms are worth knowing, because your agency and your team will use them, and you should be able to tell substance from jargon.
AEO, answer engine optimization, is about earning the direct answer. It means structuring your content so a machine can lift a clean, correct response out of it and present it in a featured snippet, a knowledge panel, or an AI Overview. Clear questions as headings, concise answers underneath, and machine-readable structure are the core of it.
GEO, generative engine optimization, goes one step further. It is about influencing whether generative AI tools cite and recommend your brand when someone asks a question in your category. Here is the uncomfortable part for anyone who thinks of SEO as an on-site task. Industry analysis in 2026 finds that roughly 85 percent of brand mentions in AI answers come from third-party pages, not from your own website. That makes earned media, analyst coverage, reviews, and your broader reputation into ranking factors. GEO is closer to public relations and brand strategy than to a meta tag. It is, by most accounts, about 80 percent strategy and 20 percent technical.
Neither of these replaces SEO. Both are the same three fundamentals aimed at a new surface. Know your audience, earn authority, remove friction. The surface is now an AI answer instead of a blue link.
What this means for you
Two numbers frame the decision. AI Overviews are reducing click-through rates by up to 58 percent on the keywords they affect, so traffic you used to count on is quietly eroding. At the same time, only about 16 percent of brands systematically track how they show up in AI answers. The gap between those figures is a window, and it will not stay open.
So the practical moves are straightforward. Start where you always should: map and segment your audience, and get specific about the jobs, pain points, and search language of each segment. Then ask your team how you show up inside AI answers on those questions today, not just where you rank. Add a measure of AI visibility, sometimes called share of model, next to your traditional ranking reports. Treat authority as an owned program, because the machines are now reading your reputation, not just your homepage. And keep the technical foundation clean, because fast, accessible, well-structured sites are the ones AI trusts enough to quote.
None of this asks you to throw out what worked. It asks you to recognize that the fundamentals you have always paid for are now being implemented on a new frontier. I optimized for AOL and Yahoo, then for Google, and now for AI engines that may never send a click. The audience research underneath all three was identical. The organizations that treat AI search as a strategy question, and not a specialist’s checklist, are the ones that will still be found when the answer no longer comes with a link.
About Agileana
Agileana builds fast, accessible, standards-based websites for government and mission-driven organizations, the kind of clean foundation that both search engines and AI systems are built to trust. If you want a clear read on how your organization shows up in this new landscape, we are happy to help you find it.