You are deciding where scarce content hours go in 2026. Classic SEO still owns commercial intent and the pages that close. Answer Engine Optimization (AEO) is the work of becoming the clear, citable source when buyers ask ChatGPT, Perplexity, Gemini, or Google AI Overviews before they hit your site. Google’s own Search Central guidance says optimizing for those AI features is still SEO. You can split the week between the two without building a second marketing department.
What you are deciding
| Workstream | What it is in practice | Wins when | Loses when |
|---|---|---|---|
| Classic SEO | Rank and convert on queries with clear commercial intent | Buyer is comparing vendors, pricing, or “best X for Y” with purchase intent | You only publish soft thought leadership with no bottom-funnel pages |
| AEO (answer / AI overview visibility) | Structured, entity-clear answers that models and AI Overviews can ground on | Category education, “what is,” compare frameworks, multi-step research | You chase “AI hacks” (chunking, fake mentions, special AI files) Google says to ignore |
| Both (Google’s frame) | Same technical eligibility + unique non-commodity content | Site is crawlable, indexed, snippet-eligible, and useful to the reader | You treat AEO as a separate product that replaces SEO |
Google’s AI features doc (updated 10 Dec 2025): there are no additional technical requirements beyond being indexed and eligible for a snippet. AI Overviews and AI Mode may use query fan-out (multiple related searches) and show a wider set of supporting links than classic Search.
fn-content has no verified atom yet for how Seed–Series B B2B SaaS teams split SEO vs AEO hours, or for AI-referral share of pipeline. fn-content tracks it as benchmark request: B2B SaaS SEO vs AEO effort split and AI-referral share. Until then, use the named public sources above and the stage table below, not an invented industry percentage.
What AEO means in practice (not a new Google product)
In operator language, AEO for B2B SaaS means:
- Answer-shaped pages. Lead with the decision, a short definition, and a comparison table. Models and AI Overviews prefer extractable structure. Google still tells you to write for humans with clear headings, not to “chunk” for machines.
- Citation readiness. Named sources, dates, and numbers that survive a check. Answer engines skip vague “industry averages” and reuse specific public citations.
- Entity clarity. Consistent product name, category, ICP, and “who it is for / not for” on the site. SparkToro (14 Aug 2026): AI Overviews, AI Mode, ChatGPT, Perplexity, and peers still learn from what they crawl on your site. Out-of-date help docs produce wrong answers.
- Supporting links over vanity rank. Google (May 2024 and May 2025): Google designed AI Overviews to show links to explore further, and supporting links can reach a wider set of sites on complex queries.
Google says you can ignore these for Google Search (AI optimization guide): llms.txt and other special AI files, rewriting copy “just for AI,” inauthentic mention campaigns, and over-focusing on structured data as an AI Overviews hack. Structured data still helps rich results. It is not a secret AEO switch.
When SEO still wins
| Query shape | Why classic SEO stays primary | What to ship |
|---|---|---|
| Pricing / plans / “cost of X” | High intent; buyers need your page, not a summary | Clear pricing page, honest bands, FAQ that matches sales |
| “[You] vs [competitor]” | Comparison shoppers convert on your framing | Fair compare tables; no keyword stuffing |
| Integration / “X for Salesforce / HubSpot” | Implementation intent | Docs that match product reality |
| Demo / trial / “alternatives to Y” | Late research with a vendor shortlist | Landing pages sales already use |
Google’s May 2025 note points to a measurement trap: AI Overview traffic can look softer on raw click volume while visits are more engaged. Score SEO pages on pipeline and time-to-value, not only GSC clicks.
When to prioritize AEO
| Situation | Why AEO hours pay | What to ship |
|---|---|---|
| New category or “what is [category]” | Assistants summarize; you want to be the grounded source | Definition + boundary + “not for” in one screen |
| Multi-step compare (“X vs Y for Series A SaaS”) | Query fan-out pulls supporting pages | Decision tables with named sources |
| Soft TOFU that used to be blog traffic | Zero-click and AI answers absorb some discovery | Keep a durable hub page; measure mentions and branded search, not only sessions |
| Sales hears “ChatGPT said…” | Wrong entity data loses deals before the demo | Fix public docs and homepage entity copy first (FounderNexus session: buyers ask AI before your site) |
SparkToro’s read of Similarweb (consumer finance/travel/beauty panel): AI recommendations correlated with higher direct visits and slightly lower search share for mentioned brands. Treat that as a signal that mentions move preference, not as your SaaS CAC model. B2B ASP and buying committees are different.
How measurement differs
| Signal | Classic SEO | AEO / AI features | Source |
|---|---|---|---|
| Impressions & clicks | GSC Performance (Web) | Same overall report plus Generative AI performance reports (impressions, pages, country, device, date) | Google Search Console, 3 Jun 2026 |
| Click quality | Time on site, conversion | Google: clicks from AI Overview SERPs are higher quality (more time on site) | Google Search Central blog, 21 May 2025 |
| Brand lift after mention | Branded query volume | Direct visits and branded search after AI mention (panel examples) | SparkToro / Similarweb, 29 Jun 2026 |
| Eligibility | Indexed + snippet-eligible | Same baseline; no extra technical gate for Overviews | Google AI features docs |
Do not buy a vendor that claims “internal Google AEO rankings.” Google’s guide: no third-party tool has access to internal ranking or AI systems.
Practical split for Seed–Series B
| Stage | Default split of content hours | Focus | Kill criteria |
|---|---|---|---|
| Seed | ~70% SEO / ~30% AEO | Pricing, ICP pages, one crisp “what we are” page AI can cite | Soft blogs with no bottom-funnel sibling |
| Series A | ~60% SEO / ~40% AEO | Compare and integration SEO; category education AEO | AEO-only calendar with no demo or pricing updates |
| Series B | ~50% SEO / ~50% AEO | Scale both; add GSC Gen AI report reviews to monthly growth | Spending on AEO “hacks” Google lists as myths |
These percentages are operator judgment for venture-scale teams with one content owner, not a survey. Align them with how often buyers mention assistants in discovery calls (FounderNexus session). If you lack a verified cohort split, track your own hours and AI-referral influenced pipeline until fn-content#20 lands.
Worked situations
Seed, one marketer, board wants “an AEO strategy.” Ship or harden pricing, one compare page, and one category definition with sources. Do not hire an AEO agency for llms.txt. Google’s guide says that file neither helps nor hurts Google Search.
Series A, organic sessions flat, demos mention ChatGPT. Add Generative AI impressions in GSC. Fix entity and docs that produce wrong answers (SparkToro’s home-base argument). Keep SEO on commercial queries. Measure branded search and direct alongside clicks.
Series B, competing on a crowded SERP for “[category] software.” Put more effort into non-commodity proof (named customers, real numbers, “not for”). Google’s May 2025 and AI optimization guides both put unique value ahead of volume. Commodity listicles lose in AI Overviews and classic results.
Sources
- Google Search Central, Optimizing your website for generative AI features on Google Search. Official guide. AEO/GEO framed as still SEO. RAG + query fan-out. Myths:
llms.txt, chunking, AI-only rewrite, inauthentic mentions. Points to Generative AI performance reports in Search Console. - Google Search Central, AI features and your website. Updated 10 Dec 2025. No extra technical requirements beyond indexed + snippet-eligible. Fan-out; wider supporting links. Higher-quality clicks note also appears here.
- Google Search Central blog, Succeeding in AI search. John Mueller, 21 May 2025. Unique non-commodity content; clicks from AI Overview SERPs are higher quality (more time on site).
- Google Search Central blog, Generative AI performance reports. 3 Jun 2026; worldwide 31 Aug 2026. Impressions, pages, countries, devices, dates for generative AI features.
- Google The Keyword, Generative AI in Search (I/O 2024). Elizabeth Reid, 14 May 2024. AI Overviews U.S. rollout; links in Overviews and greater diversity of sites on complex questions.
- SparkToro, Similarweb downstream AI visibility research. Rand Fishkin, 29 Jun 2026. Consumer panel examples: Amex +7.2%, Capital One +14.2% visit likelihood after AI mention (~7 days). Caveat: not B2B SaaS.
- SparkToro, Does your website still matter in the zero-click era?. Rand Fishkin, 14 Aug 2026. Site as home base that informs AI tools; conversion still on-site.
- FounderNexus. Founder sessions, 2026. Buyers ask AI before your site; keep bottom-funnel SEO while making category answers citation-ready. Not a survey. No closed-session numbers.
- fn-content #19. Benchmark request: B2B SaaS SEO vs AEO effort split and AI-referral share.
Related
- AI SDR vs a human SDR
- How to market an AI product without leading with AI
- Seat vs usage vs outcome pricing for AI SaaS
Founders who have rebalanced SEO and answer-engine work after buyers started asking AI before visiting the homepage will compare notes with you in a FounderNexus session.