Search used to have one job. Put a list of links in front of a person and let them choose. That job has changed. People now get answers straight from Google's AI Overviews, from ChatGPT, from Perplexity and from Gemini, often without clicking through to a single website. Which raises a fair question for anyone watching their traffic reports: is SEO still worth the effort?
The short answer is yes, and by a wider margin than the "SEO is dead" headlines suggest. But to see why, it helps to be clear on what SEO actually is, and how three newer acronyms fit around it.
Search engine optimisation (SEO) is the practice of making your website easy to find in the unpaid results of a search engine. You are not buying the placement. You are earning it by being the most relevant, credible answer to what someone typed.
Search engines decide who to show using a few core judgements, which Google sums up as E-E-A-T: experience, expertise, authoritativeness and trustworthiness. In plain terms, does this page come from someone who genuinely knows the subject, does the wider web treat them as a credible source, and can the content be trusted. Get those right and you earn traffic you do not have to keep paying for. That is the appeal of SEO and always has been. It compounds.
Two things, mostly.
First, Google started answering more questions itself. In early 2026, around 68% of US Google searches ended without a click to any website, up from roughly 60% two years earlier. When an AI Overview appears at the top of the results, click-through rates fall by close to 60%. Only about a third of searches now send a visitor onward. The search engine did not disappear. It became an answer engine as much as a directory.
Second, a share of searching moved off Google altogether and into AI assistants. Industry estimates put AI and LLM search at roughly 6% of global search volume, about triple the year before. Still small, but growing fast, and disproportionately used for the considered, research-heavy questions that matter in B2B.
That is where the two newer disciplines come in.
These are three ways of being discovered, and the good news is that they draw on the same source material.
Search engine optimisation (SEO) gets you found in the ranked results. This is the foundation, and it still governs the majority of searches.
Answer engine optimisation (AEO) gets you into the answer itself. The featured snippet, the AI Overview, the spoken response from a voice assistant. AEO is about being the source a machine quotes when it replies directly, rather than the link a person clicks.
Generative engine optimisation (GEO) gets you cited and recommended inside AI assistants such as ChatGPT, Claude, Perplexity and Gemini. You will also see this called large language model optimisation (LLMO). For most practical purposes the two describe the same goal: being the brand the model brings up.
Forget what you have read from the doom headlines heralding the death of SEO. AEO and GEO do not replace SEO. AI engines read the same web that search engines crawl. If Google cannot find, read and trust your page, an AI model usually cannot either. The technical hygiene that makes you rankable, a crawlable site, clean structure, fast pages, is the same hygiene that makes you quotable.
The credibility signals are shared too. The expertise, authority and trust that lift you up the rankings are the very things that get you pulled into an AI answer. Research points to roughly 85% of the sources AI tools cite coming from third parties rather than a brand's own website, which means being genuinely well regarded across the web, the core discipline of SEO, is what earns those mentions.
And the traffic that does still click through is worth more, not less. When someone bypasses the instant answer to visit your site, it's likely that they are further down the path and closer to a decision. Fewer visits, higher intent.
So the foundations have not moved. What has changed is the shopfront, and how many windows you now need to appear in.
The good news is that you mostly do this with one strong page, not three. Content that genuinely serves a reader tends to serve every engine. The ebook below offers some insight on this.
Answer the real question early. Lead a section with a clear, direct answer of a sentence or two, then expand. Machines and busy readers both reward it, and AI systems tend to quote the top of a page.
Structure for clarity. Use plain, question-led headings and short sections. Well-organised formats such as clear steps, comparisons and tables get quoted noticeably more often than dense blocks of text.
Put a name and a face to your expertise. Content with a credited, credible author earns more AI citations than anonymous copy. Original data, first-hand experience and a genuine point of view are hard for a competitor, or a language model, to imitate.
Keep it current. Pages that are reviewed and refreshed regularly are cited far more than ones left to age. This is one reason refreshing strong older content, like this article, beats starting from scratch.
Build credibility beyond your own site. Reviews, mentions, expert directories and third-party coverage all feed the reputation that both search engines and AI models rely on.
You cannot improve what you do not measure, and most of the tools are free to start with. Google Search Console shows which queries you appear for, where you rank and what people click. Google Analytics 4, reported through Looker Studio (the tool formerly called Google Data Studio), shows what visitors do once they arrive. Between them they answer the two questions that matter most: are you being found, and is it doing anything.
For a deeper view, paid platforms such as Semrush, Ahrefs and Moz track rankings, backlinks and competitor visibility. You will also come across authority scores like Domain Authority and Page Authority from Moz, or Trust Flow and Citation Flow from Majestic. These are useful shorthand for the strength of a site, but treat them as directional. They are estimates built by tool companies, not the way Google actually ranks pages. Rising scores are a good sign, but should not be the goal itself.
If your website and marketing already run on HubSpot, much of this sits in one place. HubSpot's SEO tools give you priority-ordered optimisation recommendations, a content strategy tool for building topic authority, and a Google Search Console integration for tracking impressions and clicks.
What is genuinely new is measuring AI visibility. Traditional tools tell you nothing about whether ChatGPT recommends you or whether you appear in a Google AI Overview. A growing set of platforms, including AI-visibility features now built into Semrush and HubSpot's AEO beta, track how often AI assistants mention or cite your brand against your competitors, sometimes called your share of model. Add the AI referral traffic now visible in your analytics, from sources like ChatGPT and Perplexity, and you can finally see whether your content is being answered with, not just ranked.
The principle holds across all of it. Watch a small number of measures that connect to real outcomes, being found, being cited and turning attention into enquiries, rather than chasing a single vanity score.
Strip away the acronyms and the picture is simple. Search rewards the businesses that are genuinely findable and genuinely credible. AI search raises the bar on both. Your job is as it always has been: be worth finding, worth answering with, and worth recommending.
That is exactly what owning your space means in practice. The new discoverability landscape is not a reason to panic about lost clicks. They are the current set of tools for making sure that when your buyers, or the machines they increasingly ask, go looking for expertise in your field, the answer they land on is you.