generative engine Optimisation is no longer a buzzword on the horizon — it is the operating reality for every site competing for visibility in search results and AI-generated answers. This week’s developments confirm that structured data, AI-influenced purchasing behaviour, and Google’s evolving spam defences are converging fast. Our team is already adjusting client strategies accordingly. Here is what matters right now and what we are doing about it.

Key Takeaways

• Schema markup has become essential infrastructure for feeding data to both traditional search engines and AI answer systems.
• AI tools are actively reshaping how B2B buyers discover, evaluate, and shortlist vendors — making brand visibility in AI outputs a commercial priority.
• Google’s latest research targets AI-generated video spam at the pattern level, signalling stricter quality enforcement across all content types.
• SEO itself has expanded beyond rankings to include visibility inside AI-generated responses, requiring a broader optimisation approach.
• Platform-level tooling and peer reviews are guiding agency and in-house teams toward more data-driven, auditable workflows.

Schema Markup Is Now Non-Negotiable for AI Visibility

Structured data is no longer just about winning rich snippets. It is the primary mechanism by which AI systems pull explicit, trustworthy facts from your pages. We now treat schema implementation as foundational infrastructure on every new build and redesign, a practice reinforced by Semrush’s detailed guide on schema markup and how to add it to your site.

For our clients, this means:

• Deploying Organisation, Product, FAQ, and Article schema as standard across all commercial and content pages.
• Validating markup quarterly to catch drift caused by CMS updates or template changes.
• Extending structured data to cover service areas, reviews, and event information — the exact entity types AI models reference when generating answers.

If your site lacks clean schema, you are effectively invisible to the generative layer sitting on top of traditional results.

AI Is Rewriting the B2B Buying Journey

A survey of over 600 US business professionals confirms what we have observed in our own analytics: AI tools now influence vendor discovery, evaluation, and shortlisting. The findings, detailed in Semrush’s research on how AI tools shape the B2B buying process, show that decision-makers increasingly rely on AI-generated summaries before ever visiting a supplier’s website.

Our response is direct. We are building topical authority content specifically designed to surface in AI outputs — long-form comparison pages, transparent pricing breakdowns, and detailed case studies with named results. If an AI tool cannot find clear, structured evidence of your expertise, it will recommend a competitor instead.

Google’s Pattern-Level Spam Detection Raises the Quality Bar

Google has published research on catching AI-generated video spam by clustering account behaviour rather than analysing individual videos. As reported in this breakdown of Google’s pattern-level AI video spam detection, the approach signals a broader enforcement philosophy: platforms will increasingly judge content producers by their overall output patterns, not isolated pieces.

For our clients producing video and multimedia content, we are tightening editorial standards and ensuring every asset has a clear purpose, original insight, and proper metadata. Bulk-produced, low-value content is a liability, full stop.

SEO Now Means Visibility Across Search and AI Answers

The definition of SEO has expanded materially. It is no longer confined to ten blue links. As outlined in Semrush’s complete guide to search engine optimisation, earning visibility now includes AI Overviews, featured snippets, and conversational AI platforms. Our technical audits, content strategies, and link-building programmes all reflect this wider scope.

Data-Driven Tooling Underpins Every Decision

Reliable tooling matters. We select platforms based on proven capability and peer validation, which is why aggregated feedback such as Semrush’s summary of over 6,300 user reviews across G2 and Capterra informs our stack decisions. Transparent, auditable data is the backbone of every recommendation we make to clients.

The thread connecting all five developments is straightforward: search visibility in 2025 belongs to sites that provide clean structured data, produce genuinely useful content, and optimise for both traditional engines and AI systems. Our team is embedding these principles into every project, from initial technical audit through to ongoing content programmes. The agencies and businesses that act now will hold the advantage as generative search matures.

Frequently Asked Questions

What is Generative Engine Optimisation and why does it matter for my website?

Generative Engine Optimisation (GEO) is the practice of structuring your site’s content and data so AI-powered search systems can accurately reference and recommend it. It matters because an increasing share of user queries are now answered by AI summaries rather than traditional search result clicks.

How do web designers use schema markup to improve AI visibility?

Web designers embed structured data — such as Organisation, Product, and FAQ schema — directly into page code so search engines and AI models can extract explicit facts. This makes the site eligible for rich results and increases the likelihood of appearing in AI-generated answers.

Why does on-page SEO now need to account for AI-generated search results?

AI Overviews and conversational AI tools pull information from well-structured, authoritative pages to compose their answers. If your on-page SEO ignores structured data and topical depth, your content is far less likely to be cited in these AI outputs.

How does Google detect AI-generated spam content?

Google’s latest research clusters account-level behaviour patterns rather than reviewing individual pieces of content in isolation. This means sites or channels that mass-produce low-quality AI content are flagged and penalised as a group, not one video or page at a time.