For twenty years, the goal of SEO was clear: rank as high as possible on Google so users would click. That goal still exists, but it is no longer the only one.
Today, a large share of searches end without a visit to any website. The user asks, reads the summary generated by artificial intelligence and moves on. In B2B, where buyers research extensively before getting in touch, this changes the rules: the brand AI cites as a reference makes the shortlist; the one that does not appear is left out without realising it.
Ranking today means two things at once: continuing to rank on Google and getting AI models to use your content as a source.
What has changed in search
AI Overviews reduce clicks
Google increasingly shows an AI-generated summary above the results. When the answer is already there, the user does not scroll down. The effect shows up in Search Console: impressions hold steady or rise, but click-through rate falls.
Not all searches are affected equally. Simple informational queries (what is, how does it work, definitions) are hit hardest. Commercial searches and those that involve comparing suppliers still generate visits, because users need to verify.
AI assistants are a new channel
ChatGPT, Gemini, Perplexity and Copilot already work as search engines for many people. They do not show ten results: they give an answer with two or three recommendations. Competition shrinks from ten positions to three mentions.
The concept of GEO emerges
GEO (Generative Engine Optimization) is the set of actions aimed at getting a company cited in AI-generated answers. It does not replace SEO: it builds on it, because the models rely largely on well-ranked content and authoritative sources.
What still works the same
Before reinventing everything, it is worth remembering what has not changed:
- Search intent rules. Understanding what the user is really looking for is still the first step.
- The website has to work. Speed, indexing, structure and the mobile version are prerequisites, not improvements.
- Authority matters. Mentions and links from relevant sources are still a signal of trust, for AI models too.
- Useful content wins. Google has rewarded content that genuinely solves problems for years. AI has accelerated that trend, not reversed it.
What to do to rank in AI as well
1. Answer specific questions directly
Models extract passages that answer a question. Content with clear sections, each answering a real query, is more likely to be cited than a long, unstructured text. Put the answer at the start of the section and the detail afterwards.
2. Provide data only you have
Models value original information. Proprietary data, benchmarks from your sector, project results or specific technical knowledge are content AI cannot generate on its own and tends to cite along with the source.
3. Strengthen authority signals
An author identified by name and job title, a complete company page, a consistent presence in industry directories and mentions in specialist media. Models cross-check sources: the more consistent the information about your company across the web, the more reliable they consider it.
4. Mark up content with structured data
Schema (Organization, Article, FAQPage, Product) helps machines understand what each thing is. It does not guarantee a citation, but it reduces ambiguity.
5. Look after commercial content, not just informational content
If AI Overviews take part of the informational traffic, pages that answer buying questions gain relative weight. Well-crafted service pages, honest comparisons and real case studies attract visitors who are close to deciding.
How to measure whether it works
Traditional metrics no longer tell the whole story. We recommend tracking these:
- Impressions versus clicks in Search Console. If impressions rise and clicks fall, AI summaries are intercepting traffic.
- Traffic from AI assistants. Visits from ChatGPT, Perplexity or Copilot appear as referrals in GA4. They are few, but usually high quality.
- Brand mentions in AI answers. Check regularly, using typical questions from your sector, whether your company appears and in what terms.
- Conversions, not just visits. In B2B, less but better-qualified traffic is preferable to more generic traffic.
Does your company appear when a buyer asks AI? Request a free visibility assessment and we will show you which sources are being cited instead of you.
What we do not recommend
Plenty of quick tactics for “getting into AI” have appeared. Most do not hold up: mass-generated content, pages created only for machines to read or texts written without human judgement. Models and search engines increasingly penalise that material, and it does not convince the person who has to sign off a purchase either.
The strategy that works is the same as always, applied more rigorously: expert content, well structured and backed by real experience.
How we work on it at Creacom
We combine technical and content SEO with visibility analysis in AI search: we check whether the company appears when a buyer asks about its sector, identify which sources are being cited instead and work on the content and authority needed to get into those answers.
Does your company appear when a buyer asks AI? Request a free visibility assessment and we will show you how Google, ChatGPT and Gemini see you.
Frequently asked questions
Is SEO still useful with artificial intelligence?
Yes. AI assistants rely largely on well-ranked content and authoritative sources, so SEO work is the foundation for appearing in their answers too.
What are AI Overviews?
They are the AI-generated summaries Google shows above the results. They answer the query directly, which reduces clicks through to websites, especially for informational searches.
How do I get ChatGPT to mention my company?
With content that answers specific questions directly, proprietary data not found in other sources, consistent authority signals across the web and structured data that makes automated reading easier.
How do you measure visibility in AI search?
By combining three signals: the relationship between impressions and clicks in Search Console, referral traffic from AI assistants in GA4 and regular checks on whether the brand appears when asking typical questions from the sector.