Where we are in the Cadence campaign
In b3 you built the content engine and drafted a Cadence pillar section. Tonight you make that content discoverable - in both the search box and the AI answer box. This is the session where "we wrote great content" turns into "and people actually find it". AI is a genuine accelerator here for mapping clusters and drafting quotable answers - and a genuine liability if you let it invent search-volume numbers. We do both: use it for the thinking, verify it for the data.
Two searches now: classic SEO + GEO 6 min live
For twenty years, "search" meant one thing: rank in the blue links so people click through. That still matters. But in 2026 a huge share of questions get answered inside an AI overview or a chat reply, where the user may never click at all. Winning discovery now means two jobs: rank in classic search, and get cited inside the AI answer.
LiveWhat GEO is, and why it is suddenly urgent3 min▶
GEO - generative engine optimization, also called AEO, answer-engine optimization - is the practice of getting your content surfaced and cited inside AI-generated answers: search overviews, chat replies, and assistant responses. It is not a replacement for SEO. It is a second front that opened fast as AI search reshaped how people find things.
- Classic SEO optimises to rank a page so a human clicks it. The win is a visit.
- GEO optimises to be the source an AI answer quotes or names. The win is a mention - sometimes without a click, which is exactly why brand and citation matter more, not less.
- They share a foundation. Clear structure, one idea per section, and answers stated up front help both. Most of your b3 structure work already earns GEO points.
Self-studyClassic SEO vs GEO at a glance2 min read▶
| Classic SEO | GEO / answer-engine optimization |
|---|---|
| Goal: rank in the results list | Goal: be cited inside the AI answer |
| Win = a click to your page | Win = a mention, often with no click |
| Rewards keywords, links, page authority | Rewards clear answers, structured facts, entity clarity |
| Measured in rankings and organic traffic | Measured in citations and share of AI answers |
You do not pick one. In 2026 a strong page earns both - and the structure that wins GEO is mostly the reader-first structure you already learned in b3.
Topic clusters + intent 6 min live
Ranking is not about stuffing one page with keywords. It is about owning a topic - a pillar page on the big theme, surrounded by a cluster of supporting pages that each answer a specific question, all linked together. AI is excellent at helping you map that cluster around search intent, not just words. Just do not let it invent the numbers.
LiveMap a pillar + cluster around intent, not keywords3 min▶
A keyword tells you the words someone typed. Intent tells you what they actually wanted - to learn, to compare, or to buy. The same phrase can carry different intent, and matching it is what earns the ranking and the click.
- Pillar page covers the broad topic ("AI meeting notes") and links out to every supporting page.
- Cluster pages each answer one specific intent ("how to share meeting action items", "AI note-taker vs manual notes") and link back to the pillar.
- Match the intent. Informational queries want a clear explainer; comparison queries want an honest comparison; buying queries want the offer. AI helps you sort a keyword list into these buckets fast.
On-page basics still apply: a descriptive title, clean heading hierarchy, structured facts, and internal links between the pillar and its cluster so search engines see the topic as one authoritative set.
Self-studyGEO tactics: how to earn the citation2 min read▶
| GEO tactic | Why it works |
|---|---|
| State the answer clearly, up top | Answer engines lift the first clear, direct response - bury it and you lose the citation |
| Structure facts (lists, tables, steps) | Structured facts are easy for a model to extract and quote accurately |
| Write quotable definitions | A crisp one-sentence definition is exactly what an AI answer pulls verbatim |
| Entity clarity - name things plainly | Clear "what it is / who it is for" helps the model attribute the answer to you |
| Keep facts current and sourced | Models favour fresh, consistent, verifiable facts over vague marketing prose |
The volumes that did not exist. A marketer asked a chatbot for "monthly search volumes" for a keyword list and pasted them straight into a content plan. Half the numbers were invented - the model does not have live search data and will confidently make it up. Leadership approved a quarter of work on fiction. Use AI to group and prioritise keywords; use a real keyword tool for the actual numbers.
Programmatic scale, carefully 4 min live
Once you have a cluster template that works, AI makes it tempting to spin up hundreds of pages at once - programmatic SEO. Done well, it can own a whole category of long-tail queries. Done carelessly, it produces thin, near-duplicate pages that search engines penalise. The difference is a quality gate.
LiveScale the pattern, gate the quality2 min▶
Programmatic SEO means generating many pages from one template plus a data set - one page per city, per use case, per integration. AI can draft them at scale. That is the opportunity and the trap.
- Each page must earn its place. Real, specific value per page - not the same paragraph with the noun swapped. Thin, templated sameness is exactly what quality updates target.
- Dedup ruthlessly. If two pages say essentially the same thing, merge them. Near-duplicates compete with each other and drag the whole set down.
- Gate before publish. A human quality check and a fact-check on every generated page. Scale without a gate is how you turn AI leverage into an SEO penalty and a pile of sludge.
Build a Cadence topic cluster ★ 13 min · everyone builds
Turn the Cadence pillar theme into a real topic cluster - one pillar plus five or six supporting pages, each mapped to a specific search intent. AI proposes; you validate the intent and kill anything that does not fit a real reader.
Open any chat AI. Give it Cadence's positioning from b2 and your pillar theme from b3 (for example, "AI meeting notes for busy teams").
Ask it to propose one pillar page plus five or six supporting cluster pages, and for each, to state the search intent (informational, comparison, or buying) and the specific question that page answers.
Validate the intent yourself. Read each proposed page: is that a real question a Cadence buyer would ask? Cut the ones that are keyword-shaped but human-empty. Trust your knowledge of the audience over the model's guess.
Ask for the internal-linking map: which cluster pages link to the pillar, and how the pillar links back. This is what makes search engines see one authoritative topic.
Do not accept any search-volume or difficulty numbers the model offers - it does not have that data. Mark those as "verify in a real tool" and move on.
Rewrite a page to be answer-engine friendly ★ 8 min · everyone builds
Take one section of a Cadence page and rewrite it so an AI answer would happily quote it: a clear answer up front, structured facts, and a crisp, quotable definition. This is GEO in one concrete edit.
Pick a page section that answers a real question - say, "What is an AI note-taker?" or "How do AI meeting notes work?" Use your b3 draft if you have one.
Rewrite the opening so the answer comes first - one or two direct sentences before any context. If an engine read only the first line, it should already have the answer.
Add a quotable one-sentence definition and structure the supporting facts as a short list or steps, so a model can lift them cleanly and accurately.
Check entity clarity: is it obvious what Cadence is and who it is for? Make sure your name and category come through, since a citation may arrive with no click.
Fact-check every claim in the rewrite - GEO does not suspend rail one. An answer engine quoting your fabricated stat is your fabricated stat, at scale.
The answer-first paragraph that started getting cited. A team moved one buried definition to the top of a page and added a one-line "X is..." sentence. Within weeks that line began showing up, quoted, inside AI answers - and their brand name rode along with it. No new content, just structure that made the page easy to quote. GEO is often an edit, not a rewrite from scratch.
Try it yourself - this week ◐ 30-45 min total
- Finish the Cadence topic cluster: one pillar, five or six validated cluster pages, each with a confirmed intent and a link back to the pillar.
- Take the volume/difficulty items you marked "[verify in tool]" and check two or three in a real keyword tool. Note how far the model's guesses would have been off.
- Run the answer-engine rewrite on one live page of your own. Move the answer to the top, add a quotable definition, structure the facts.
- Search two of your key questions in an AI assistant today. Are you cited? Is a competitor? That gap is your GEO to-do list.
- Optional: sketch one programmatic page type you could scale later - and write the quality gate you would put on it before you publish a single one.
Official sources covered
Taught from public AI-marketing curricula and vendor documentation. Certificates and paid tooling stay with the official programs. This session covers:
Three questions before you go 🎯 ◐ 90 seconds
1 · What is the difference between classic SEO and GEO?
Classic SEO wins the click by ranking in the results; GEO (answer-engine optimization) wins the mention by being what an AI answer quotes. In 2026 you optimise for both, and they share a structural foundation.
2 · What is a topic cluster, and what should each page target?
A cluster is a pillar page surrounded by supporting pages, each answering one specific question mapped to intent (learn, compare, buy) - not just a keyword. Internal links tie them into one authoritative topic.
3 · The AI gives you neat monthly search volumes for your keyword list. What do you do?
Chat models do not have live search-volume data and will fabricate confident numbers. Use AI to group and prioritise keywords; get the actual volumes and difficulty from a real SEO tool.