Weekly AISEO + GEO Roundup (week of 27 Jul 2026)

Google just published official guidance on “optimising for generative AI features” (AI Overviews and AI Mode). The headline: there’s no secret GEO playbook - showing up in AI answers is still driven by excellent SEO fundamentals, clear writing, and genuinely useful, non‑commodity content. This week’s roundup breaks down the signals, what they mean for SMEs, and a practical checklist you can action in 60-90 minutes.

Weekly AISEO + GEO Roundup (week of 27 Jul 2026)

1) What people are talking about (signals)

Google is now treating AI Overviews / AI Mode as a mainstream Search surface with mainstream guidance: website-owner best practice, not a separate “GEO checklist”.

Why this matters: it’s evidence that (for Google, at least) “GEO” isn’t a new discipline - it’s a new presentation layer for the same underlying ranking and quality systems.

Google is also being explicit that there are no extra requirements to appear in AI Overviews/AI Mode. That implicitly pushes back on a wave of “AI-specific tactics”:

• “special schema just for AI”
• llms.txt as a requirement
• “chunking” as a magic bullet
• other “AI hack” checklists with no measurable outcome

The practical takeaway: if you can’t tie an activity to a measurable output (qualified visits, leads, assisted conversions, citations/mentions), it’s probably distraction.

The SEO community consensus is doubling down on what many SMEs have avoided because it takes effort: unique, proof-heavy content.

“Non‑commodity” content is anything that can’t be reproduced by summarising the top 10 results. Think:

• first‑hand experience (what you’ve seen in the field)
• mini experiments and real outcomes
• screenshots, examples, before/after
• proprietary data (even small datasets)
• clear opinions backed by reasoning and evidence

This aligns neatly with Google’s public framing: usefulness, originality, clarity.

CTR impact discussions haven’t gone away - they’ve become more nuanced.

• Yes: AI answers can reduce clicks on informational queries.
• Also yes: being cited/referenced can still drive discovery, branded demand, and downstream conversion.

For SMEs, this is especially important: you may not win by chasing raw sessions. You win by increasing recognition and trust so more of the right people search your brand later, or convert when they’re ready.

Reddit keeps showing up as a practical signal in marketer communities - partly for brand perception, partly because some answer engines appear to source “human experience” from discussion platforms.

You don’t need to “do Reddit marketing” by default - but you should treat it as a place where your category narrative can be shaped (accurately or not).

The emerging framing is:

• For Google: “GEO is still SEO.”
• For multi-engine visibility (ChatGPT/Perplexity/etc.): you still need distribution, corroboration, entity clarity, and proof.

In other words: build the best possible site experience and content base, then ensure the rest of the web agrees with who you are and what you do.

2) What it means in practice (implications for SMEs)

For Google AI features, the playbook looks like:

• clear intent matching (answer the query quickly)
• helpful structure (headings that map to sub-questions)
• original value (proof, examples, experience)
• strong internal linking to consolidate authority

If you’re an SME, this is good news: you don’t need a new technical project to be eligible. You need clarity and credibility.

If someone is selling a GEO technique, ask:

• What outcome changes if we do this?
• How would we know within 2-4 weeks?
• What would we stop doing if it doesn’t work?

If you can’t answer those, it’s likely not worth the time.

For informational queries, sessions alone are increasingly incomplete. Add:

• answer presence / citations (are you being referenced?)
• brand recall (are branded searches increasing over time?)
• assisted conversions (did someone first discover you via informational content, then convert later?)

AI systems need to understand:

• who you are
• what you do
• where you operate
• why you’re credible
• how your authors are connected to the business

That means consistency across your site, structured data, and corroborating third-party references.

“Quotable” content is easier to cite and reuse:

• tight sections
• crisp headings
• direct answers
• tables where useful
• clear definitions and examples

This is a writing discipline, not a dev sprint.

3) Actions for clients this week (concrete)

Below is a practical checklist you can run without a full site rebuild.

Pick your 20 highest-impression informational URLs.

For each page:

1. Rewrite the intro to answer the core query in 40-60 words.
2. Immediately follow with supporting proof:

Rule of thumb: answer first, then earn trust.

• Validate Organisation schema and ensure sameAs links point to official, authoritative profiles (company social profiles, reputable directories, partner pages).
• Add Person schema to author pages, and connect authors to the organisation and to external profiles where appropriate.

This is less about “schema magic” and more about consistent identity.

If you publish advice, explain:

• how you create content
• how you test claims
• how often you update
• who reviews or edits

This increases trust for both humans and systems.

For each priority topic, create one URL that becomes your source of truth with:

• mini experiments
• screenshots
• outcomes
• updated findings

Then link to it internally from every related article.

Pick 3 high-performing posts and add an FAQ section:

• 6-10 questions you genuinely hear in sales/support
• concise, specific answers
• avoid fluff and generic “it depends” where possible

Ensure your key facts are consistent on 3-5 credible third-party sources:

• industry directories
• partner pages
• press/PR pages
• major social profiles

This is “entity verification by consensus.”

Pick 10 “money query” prompts and test weekly across:

• Google AI features (Overviews/AI Mode)
• one LLM (ChatGPT or Perplexity)

Log:

• whether you’re cited
• which URL is cited
• which competitor appears instead

Consistency beats complexity here.

Identify thin pages that restate the obvious. Either:

• add first‑hand evidence, or
• consolidate them into a stronger canonical page

Task list (copy/paste)

• ☐ Export your top 20 informational URLs (highest impressions) from Search Console
• ☐ Rewrite each intro to answer the query in 40-60 words
• ☐ Add one “proof block” per page (example, screenshot, mini case, data point)
• ☐ Validate Organisation schema + add sameAs links to authoritative profiles
• ☐ Ensure each article has an author page; add Person schema and link author → organisation
• ☐ Publish/refresh About + Editorial/Methodology pages (with update cadence)
• ☐ Create one evidence hub page for your #1 priority topic and link to it from related posts
• ☐ Add a 6-10 question FAQ section to 3 high-performing posts
• ☐ Check 3-5 third-party profiles for consistent brand facts (name, location, offering, leadership)
• ☐ Set up weekly “answer presence” tracking for 10 prompts (Google AI + one LLM)
• ☐ Merge or upgrade 5 thin/commodity pages

4) Three article topics to publish next (based on this week’s signals)

1. Google’s AI Overviews Optimisation Guide: What It Actually Changes (and What It Doesn’t)
2. Entity Clarity for AI Search: The 10 Signals That Make Brands “Citable”
3. Visibility Without Clicks: How to Report AI Search Impact to Clients in 2026

Glossary (updated)

• AI Mode (Google Search): a more interactive, agentic AI search experience (distinct from AI Overviews).
• Non‑commodity content: content that adds unique value (original experience, proprietary data, first-hand testing) and can’t be replaced by generic summaries.
• Answer presence / citation tracking: monitoring whether your brand/URLs are referenced inside AI-generated answers, not just ranked in blue links.