Published by ELK Digital — April 2026

Every few years, someone announces the death of SEO. It happened with Panda, with mobile-first indexing, with voice search, with BERT. Each time, SEO didn’t die — it changed shape. What’s happening in 2026 is different in degree, not in kind. The blue-link web is no longer the finish line. For a growing share of queries, it’s a footnote underneath an AI-generated answer.

So if you run a business and you’re wondering whether your SEO agency is quietly selling you yesterday’s playbook, the honest answer is: probably partly yes, and partly no. Here’s what’s actually changed, what hasn’t, and what you should be asking for in 2026.

The short version

Search is no longer one discipline. It’s three, working together:

  • SEO — getting found in traditional search results. Still matters. Still feeds everything else.
  • AEO (Answer Engine Optimisation) — getting your content used as the direct answer in featured snippets, voice results and People Also Ask boxes.
  • GEO (Generative Engine Optimisation) — getting your brand and content cited, summarised or recommended by AI tools like Google’s AI Overviews, Gemini, ChatGPT and Perplexity.

If your current strategy only covers the first one, you’re optimising for a shrinking window of user behaviour. If someone tells you SEO is dead and you should only do GEO, they’re selling you a different kind of nonsense. The three disciplines share a spine — good content, clean technical foundations, genuine authority — and the work is in how you adapt each layer for how people actually search now.

What actually changed

Until recently, the pitch was simple. You typed a query, Google showed you ten blue links, you clicked one, you landed on a site, the site got a visitor, the visitor (hopefully) became a customer.

That loop is still intact for a lot of transactional and navigational queries. If someone searches “book flights to Lisbon” or “JD Sports Basingstoke opening hours,” they’re going to click something.

Where it’s broken is the informational layer. Someone searches “what’s the difference between a Roth IRA and a traditional IRA” or “how do I clean a cast iron pan” and Google now often answers inside the results page itself. The user gets what they came for without clicking. Meanwhile, a growing segment of users have stopped starting at Google altogether — they open ChatGPT, Perplexity or Gemini and ask in full sentences.

The practical consequence: if your content strategy is built around ranking for informational keywords to drive traffic that eventually converts, you’re losing the top of the funnel even when you’re winning the rankings. The impression is going to the AI summary. The citation — if you get one — is the new click.

SEO: still the foundation, but narrower in scope

Technical SEO, site architecture, Core Web Vitals, internal linking, high-quality backlinks — none of this has stopped mattering. AI-generated answers are built on top of crawled and indexed content. If search engines can’t crawl your site, render it, or trust it, you’re invisible to the AI layer too.

What’s changed is the ROI framing. Ranking #1 for a high-volume informational keyword in 2020 meant meaningful traffic. In 2026, it might mean your paragraph gets used as the source for an AI Overview that nobody clicks through. The rank is still there. The traffic isn’t always.

This is why pure “ranking reports” as the headline KPI are quietly becoming useless. What matters is whether the ranking actually delivers something — a visit, a citation, a cited mention in an AI answer, a branded search downstream.

AEO: winning the answer, not just the rank

AEO is the discipline of structuring content so search engines can lift a clean answer out of it. This has been around since featured snippets launched, but it’s now the bridge between classical SEO and the AI layer.

The practical work:

  • Lead with the answer. If someone asks a question in your H2, the next 40–60 words should answer it directly, before any context, storytelling or setup. You can add nuance afterwards. The summary goes first.
  • Use question-based headers. H2s and H3s written as real questions (“How long does a WordPress site take to build?”) outperform keyword-stuffed headers for both featured snippets and AI summarisation.
  • Structured data isn’t optional any more. Properly implemented FAQPage, Article, HowTo, Product and Organization schema in JSON-LD is now the baseline. It tells Google (and by extension, the AI systems pulling from Google’s index) what each block of content actually is.
  • Keep answers self-contained. An AI system pulling a snippet wants a paragraph that makes sense on its own, without “as mentioned above” references to context the AI didn’t lift.

Done well, AEO earns you the featured snippet position — and a featured snippet is the single biggest predictor of being cited in an AI Overview. Google tends to pull from what it already trusts at the top of its own results.

GEO: the new discipline

GEO — Generative Engine Optimisation — is the work of making your brand and content discoverable, citable and recommendable inside generative AI systems. It’s the newest layer, the least understood, and the one most agencies are still figuring out in public.

There are two sides to it.

On-page GEO is about making individual pages easy for an AI to understand, extract and attribute. That means:

  • Clear, declarative language (AI systems summarise clarity; they mangle clever)
  • Explicit entity references (use the full brand name, product name, person’s name — don’t rely on pronouns)
  • Factual density: concrete numbers, dates, specifications, named sources
  • Semantic HTML and clean heading hierarchy, so the AI can parse the structure
  • JSON-LD schema that declares what the page is and how it relates to other entities

Off-page GEO is the harder, slower half — building the authority signals that make an AI decide your brand is worth citing in the first place. AI systems don’t just lift from the top-ranked page. They weigh mentions across the web, cross-reference entities against knowledge graphs, and lean heavily on sources that are corroborated in multiple places.

In practice that means:

  • Getting mentioned (not just linked) in credible publications, industry media and trade press
  • Making sure your brand exists cleanly in Wikipedia, Wikidata and Google’s Knowledge Graph where relevant
  • Being active, genuinely, on the platforms AI models are trained on and pull from — Reddit, Stack Overflow, LinkedIn, YouTube transcripts, Quora
  • Building a consistent entity footprint: same name, same descriptions, same positioning across every profile, directory and bio

This last point is where a lot of businesses trip up. If your company is “Smith Digital” on your website, “Smith Digital Ltd” on Companies House, “Smith Digital Services” on LinkedIn and “John Smith Digital” on your Google Business Profile, you’re fragmenting your own entity. Fix that first.

Where Gemini, ChatGPT and the rest fit in

It’s worth being specific about the ecosystem, because “AI search” is doing a lot of heavy lifting as a phrase.

  • Google AI Overviews and Gemini pull heavily from Google’s own index, so strong classical SEO plus clean schema plus featured-snippet-grade answers remains the most direct route in.
  • ChatGPT’s search blends OpenAI’s own index with Bing results and a weighting towards sources that appear frequently in its training data — which is why Reddit, Wikipedia and long-established publications are over-represented in citations.
  • Perplexity is the most citation-driven of the three; it genuinely shows its sources and rewards clean, factual, well-structured pages that answer a query precisely.
  • Bing, Copilot and DuckDuckGo’s AI features all use variations on this pattern — indexed content, weighted by authority signals, summarised into an answer.

The good news: you don’t optimise differently for each one. The work is largely the same. Be clear, be cited, be structured, be trustworthy. The platforms that reward this now will be joined by others using the same logic.

What you should actually do in 2026

If you strip out the jargon, the practical changes to your strategy look like this:

  1. Audit what you already rank for and ask what it’s worth. If you’re ranking for informational queries that now get answered by AI Overviews, accept that the click-through rate has dropped and plan accordingly — either shift those pages towards brand-building and citation value, or reallocate effort to commercial-intent pages where users still click.
  1. Rewrite your highest-traffic pages to lead with a direct answer. First 40–60 words should answer the headline question on its own. Everything else supports it.
  1. Get your structured data right. Not “we added some schema three years ago” — a proper audit, proper JSON-LD, proper entity linking between your Organization, your content and your products or services.
  1. Build a question-first content plan. Instead of keyword lists, work from real questions your customers ask. Mine support tickets, sales calls, Reddit threads, your own inbox. These are the prompts people type into AI systems.
  1. Treat brand mentions as SEO currency. A mention in a trade publication with no backlink is now worth real money. It feeds entity recognition. Invest in digital PR alongside (not instead of) link building.
  1. Get your entity footprint consistent. Same name, same description, same positioning, everywhere. Wikipedia, Wikidata, LinkedIn, Google Business Profile, industry directories, your own site’s Organization schema. Make the AI’s job of recognising you easy.
  1. Track new KPIs. Traditional rankings and organic sessions still matter. Add to them: share of voice in AI Overviews for your priority queries, citation frequency in ChatGPT and Perplexity, branded search volume (a strong proxy for AI exposure), and assisted conversions from AI-driven sessions.

The honest summary

Traditional SEO isn’t dead. It’s been absorbed. The technical rigour, the authority signals, the content quality — all of it still runs the show. What’s dead is the assumption that ranking is the finish line and a click is guaranteed.

The businesses that will do well in 2026 and beyond aren’t the ones chasing the new acronym. They’re the ones who’ve understood that SEO, AEO and GEO are three facets of the same job: being the clearest, most credible, most structured answer to the question a customer is asking — whether that customer finds you via a blue link, a snippet, a voice assistant or a line of AI-generated prose with your name cited underneath.

The playbook isn’t dead. It’s just longer.

ELK Digital is a Hampshire-based digital consultancy specialising in ecommerce, WordPress and Shopify development, and AI-era SEO strategy. If you’d like a conversation about where your site stands in the new search landscape, get in touch — or if you’d rather know where you stand measured than guessed, that’s what the AI-Search Readiness Audit does: £495, fixed, inside a week.