Search has quietly stopped being a list of blue links. Somewhere between the first AI Overview and the rise of agentic browsing, the job of SEO changed from clicks to getting cited, thanks to an AI-powered SEO strategy.
This guide by Brand Mender pulls together what the top platforms actually building AI search have to say about it in their 2026 guides.
We shall explore insights offered by:
- Adobe
- Salesforce
- HubSpot
- And More
From Ranking Pages to Earning Citations
For two decades, SEO ran on a simple loop:
- Rank, get clicked, and convert.
The good and the bad news – That loop is fraying!!
Adobe’s analysis of the 2026 search landscape argues that visibility now depends less on page position and more on whether a brand is cited inside an AI-generated answer, and it isn’t a small trend.
Adobe’s own research found that generative-AI-driven referral traffic in the US grew more than tenfold between July 2024 and February 2025, and that visitors arriving via AI tools browsed noticeably more pages and bounced far less than traffic from traditional channels.
In other words, the people AI sends you aren’t just curious clickers, but they arrive pre-qualified.
HubSpot’s research tells a more tempered version of the same story of AI-powered SEO strategy. HubSpot notes that roughly 80% of desktop searches still happen on traditional search engines, with AI tools accounting for a small single-digit share. The honest takeaway sits between the two extremes:
- AI search isn’t replacing traditional search, but it is becoming a real, commercially meaningful discovery layer that brands can no longer treat as optional or experimental.
This isn’t really a story about traffic volume. It’s a story about where the decision gets made.
Even when someone eventually clicks through from a traditional search result, they may have already formed an opinion based on an AI Overview, a ChatGPT answer, or a Perplexity summary they saw earlier that day. Losing visibility in that earlier moment means losing influence over the decision, even if your click-through numbers look fine on paper.

Is SEO Dead? Short Answer – No!
It’s tempting to read AI Overviews and zero-click search as a eulogy for SEO in 2026. The data doesn’t support that. HubSpot points to SparkToro founder Rand Fishkin’s zero-click research, which found that while the percentage of clicks going to the open web has dropped, the absolute number of people clicking away from Google has stayed roughly flat, because total search volume keeps growing.
Fewer clicks per search, more searches overall, roughly the same number of visitors.
What’s actually changed is the metric that matters:
- Traditional SEO measured rankings, clicks, and organic sessions; AI search optimisation measures citation frequency, share of model (how often you show up relative to competitors across AI platforms), and AI-referral traffic.
Google’s own developer documentation states plainly that, from Search’s perspective, optimising for AI Overviews and AI Mode is not a separate discipline from SEO. It’s still SEO, because these features are built on the same core ranking and quality systems Google has always used.
That distinction matters more than it sounds for SEO in 2026. Google is telling marketers not to chase a parallel set of “AEO hacks,” while Adobe and HubSpot are building entire product categories around the idea that AI visibility needs its own measurement stack.
Both things can be true. Hence:
- The underlying content principles haven’t changed much, but the reporting and prioritisation absolutely need to.
What Google Actually Says to Ignore
This is the most useful part of Google’s guidance for future SEO, because it directly contradicts a lot of what’s being sold as AI SEO hacks by many people on online forums.
Google explicitly lists tactics we can skip now:
- txt files – Google Search doesn’t read them. Creating one won’t hurt you, but it won’t help your Google visibility either.
- Obsessive “chunking” – You don’t need to fragment content into bite-sized blocks purely for AI parsing. Google’s systems already understand multiple topics within a single page.
- Rewriting for AI instead of humans – Google’s models handle synonyms and intent matching well enough that keyword-stuffing every possible phrasing is wasted effort.
- Chasing inauthentic brand mentions – Planting your name across forums and blogs to manufacture “signals” doesn’t move the needle and risks running into spam policies.
- Treating structured data as mandatory – It helps with rich results, but it isn’t a requirement for appearing in AI Overviews.
What Actually Moves the Needle
Setting aside the myths, a consistent picture emerges across all four platforms about what genuinely helps.
- Write content that only you could write. The famous E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness) is to be followed. A generic 7 tips listicle is exactly the kind of content an AI model can synthesise without ever needing to cite you. A first-hand account of a campaign you actually ran, results included, is much harder to replace.
- Make authorship visible and specific. HubSpot highlights that Google explicitly rewards clear bylines linked to author profiles with demonstrated subject-matter background.
- Cover a topic completely, in one place. A comprehensive core page supported by linked deep dives consistently correlates with better AI citation rates, because it demonstrates depth rather than a single isolated answer. This also plays into what HubSpot calls “entity optimisation”, which means being consistently and clearly described across your own site and third-party sources so AI knowledge graphs can confidently associate your brand with a topic.
- Don’t ignore the technical basics. None of the above matters if a page isn’t indexed. Google is unambiguous that a page must meet standard technical requirements to be eligible for any AI feature — it isn’t a separate on-ramp.
Optimising Content for AI Retrieval
- Creating high-quality content is only part of the equation now. To improve the likelihood of being cited by AI systems, brands should also make their information easy to retrieve and interpret.
- Pages should answer the primary question early, use clear headings that reflect user intent, and organise related ideas into logical sections without unnecessary repetition.
- Supporting claims with original research, case studies, statistics, or expert commentary strengthens credibility and provides unique value that AI models cannot easily reproduce.
- Consistent internal linking between pillar pages and supporting resources also helps search engines understand topical relationships while guiding users to deeper information.
- Rather than producing dozens of similar articles targeting slight keyword variations, focus on building comprehensive resources that satisfy multiple related queries within a single authoritative page.
- This approach improves both traditional search performance and the chances of being referenced in AI-generated responses across different platforms.
The Search Experience Itself Is Changing
It’s worth understanding where Google is taking Search, because it reshapes what visibility will even mean going forward.
At its 2026 I/O event, Google announced that AI Mode has surpassed one billion monthly users, with query volume more than doubling every quarter since launch.
Google also introduced a redesigned AI-powered search box—its biggest interface change in over 25 years—along with multimodal input for text, images, video, files, and even Chrome tabs, as well as Search agents.
These information agents run continuously in the background, monitoring the web for updates relevant to a standing query. For example, they can monitor apartment listings matching your criteria or a sneaker drop from a favourite athlete. They proactively notify the user, with the ability to take action like booking on their behalf.
Google is also piloting agentic calling, where users can ask Google to call a local business for them, and agentic coding that builds custom trackers and mini-apps on the fly from a single question.
This matters for one blunt reason:
- If agents are completing tasks autonomously based on content they’ve read on your site, ranking well stops being enough.
Rather than chasing every new acronym, a workable approach looks like this:
- Keep your technical SEO clean – Indexability, crawlability, page experience, and structured data remain the entry ticket, not the differentiator.
- Build fewer, deeper pieces of content – Comprehensive, pillar-style coverage of a topic outperforms thin content spread across many pages, both for rankings and for AI citation.
- Put a name and a track record behind your content – Authorship with demonstrable, relevant experience is one of the few signals AI genuinely can’t fabricate on your behalf.
- Measure citations and share of model, not just clicks – If your reporting stack only tracks organic sessions, you’re blind to a growing share of how buyers actually form opinions.
- Watch the agentic layer – As Search agents and AI shopping/booking tools mature, structure your product, pricing, and specification content so it’s unambiguous enough for a machine to act on correctly.
Conclusion
Being helpful, specific, and verifiably credible was always core to good SEO. It has simply become the entire game now that answers are synthesised rather than merely linked.
The brands that treat AI visibility as infrastructure will be the ones still getting cited and chosen, long after the wave of AEO hacks has been quietly ignored by the algorithms.
Frequently Asked Questions
- Is AEO (Answer Engine Optimisation) a different discipline from SEO?Depends who you ask, and it's worth knowing both answers. Adobe and HubSpot treat AEO/GEO as an emerging specialisation with its own metrics (citation frequency, share of model) and tooling.Google takes the opposite stance in its own guidance, stating that optimising for its generative AI features is SEO, not a separate practice, since both are powered by the same core ranking systems. In practice, use AEO as a useful lens for measurement and reporting, but don't treat it as a wholly separate playbook from good SEO fundamentals.
- Do I need schema markup and llms.txt files to show up in AI Overviews?No, at least not for Google. Google has explicitly said that llms.txt files are ignored by its systems. While structured data is helpful for rich results, it is not required for AI feature eligibility. Treat schema as good practice for traditional SEO gains, not as an AI-specific shortcut.
- What's the single highest-leverage change a content team can make right now?Across every source referenced here, the recurring theme is originality with visible expertise.
- Named authors with real credentials
- First-hand experience
- And content that goes beyond what a model could already synthesise from existing sources.

















