SEO &amp; GEO Dual-Engine Playbook

Key takeaways

Start from BOFU — do high-intent terms (pricing, comparisons) first; they convert highest, then work up to educational content.
Make provenance visible — use founder voice, first-hand examples, dates, scope, and primary sources so claims can be checked.
Structure supports extraction — Key-Stats tables, direct answers, and visible FAQs make passages easier to parse, without guaranteeing citation.
IndexNow is discovery infrastructure — notify participating search engines when content changes, then verify crawl, index, and citation separately.
Research Google page 1 before writing — the top 5-6 results already give you 80% of the structure; improve it by replacing outdated data and adding your own product GIFs.
Refresh winners annually — update product and competitor screenshots in every high-ranking article each year; freshness keeps compounding rankings alive.
Audit before you publish — get your site to 90+ on an SEO audit tool before producing content, or keywords will not rank.

What this is

Google ranking and AI-search citation are two different games with one content system. SEO rewards backlinks + keywords; GEO rewards structured data + E-E-A-T authenticity. This playbook runs both engines from the same content — so a single page earns Google rankings and shows up inside ChatGPT/Perplexity/Claude answers.

⚠️ Product type note: This playbook is optimized for dev tools / OSS / B2B SaaS. If your product is a 2C consumer app, education platform, or game, key parameters differ — lower keyword volume threshold (50 vs 300), YMYL E-E-A-T requirements, and different cold-start channels (Reddit communities / 小红书 / TikTok / Quora instead of PH/GitHub/HN). See the 2c-adaptation guide in the skill's references/ folder.

Key results

Months to 60K stars (AFFiNE, SEO-driven)	24
Startup consultations behind the system	150+
Core principles	5
Engines optimized from one content set	Google + 4 AI
Traffic growth in first 3 months of SEO (AFFiNE, self-reported)	~25x
Monthly impressions from SEO alone, zero paid (self-reported)	~400-500K
Proven content cadence	4 articles/week

The 5 core principles of SEO + GEO

1
1 · Start from BOFU, work upward
Build the highest-intent, highest-converting keywords first (pricing, comparisons, alternatives), then expand up the funnel to educational content. Revenue-adjacent pages compound fastest.
2
2 · Make provenance visible
Use the founder's voice, real numbers, first-hand stories, dates, and primary sources so readers and systems can verify each important claim.
3
3 · Structure supports extraction
Use Key-Stats tables, visible FAQs, and comparison matrices where they help readers. Structured, answer-first content is easier to extract, but markup alone does not guarantee citation.
4
4 · IndexNow supports discovery
Use IndexNow to notify participating search engines when content changes. Treat discovery, crawling, indexing, ranking, and AI citation as separate outcomes.
5
5 · Comparison pages are SEO gold mines
Build one dedicated page per competitor ("X vs Y", "X alternatives") to capture decision-stage users — these are the highest-ROI pages for both SEO and GEO.

SEO vs GEO — two engines, one content system

Optimize for both at once: the same page should rank on Google and be quotable by AI search.

Dimension	SEO (traditional)	GEO (AI search)
Goal	Google / Bing rankings	Cited inside AI answers
Core signal	Backlinks + keyword match	Structured data + E-E-A-T
Content format	H2/H3 + long-tail coverage	Direct answers + tables + FAQ schema
Technical	XML sitemap + Core Web Vitals	IndexNow + AI-crawler-open robots.txt

Anti-patterns (what kills SEO + GEO)

AI-sounding generic copy — no E-E-A-T signal, no citation.
Cherry-picked comparison rows — erodes trust and costs you AI citations.
Starting from TOFU educational content — slow, low-converting; do BOFU first.
Treating schema as a ranking or citation guarantee instead of a machine-readable description of visible content.

FAQ

What is GEO and how is it different from SEO?

GEO (Generative Engine Optimization) is optimizing to be cited inside AI-search answers (ChatGPT, Perplexity, Claude, Gemini), where the signals are structured data + E-E-A-T authenticity. SEO targets Google/Bing rankings via backlinks + keywords. This playbook wins both from one content set.

What's the single highest-ROI thing I can do?

Build comparison pages — one dedicated "X vs Y" or "X alternatives" page per competitor. They capture decision-stage users and are equally strong for Google ranking and AI citation.

Do I really need IndexNow?

IndexNow is useful when you publish or update URLs because it notifies participating search engines of the change. It does not guarantee crawling, indexing, ranking, or AI citation, so verify each stage separately.

Who built this?

Iris Wei (生姜) — former cofounder/COO of AFFiNE (60K GitHub stars in 24 months, largely SEO-driven) and advisor to 150+ startups on search and AI-search growth.

How fast can zero-budget SEO actually work?

At AFFiNE (2022-2023, self-reported), traffic grew ~25x within 3 months of starting SEO and reached ~400-500K monthly impressions with zero paid spend — on a cadence of 4 articles/week (~20/month), each ~1,500 words with images, GIFs and tables, refreshed annually.

What does SEO compounding look like in practice?

One AFFiNE article held Google positions 3-4 for 'open source notion alternative'. When Notion banned Russian accounts, searches spiked and that single old article automatically captured a wave of Russian users — high-authority pages catch event-driven traffic for free.

Related playbooks

Growth Finder — diagnose the next growth constraint→ Open-Source Marketing — 0 to 60K GitHub Stars→ B2B SaaS Growth — PLG vs SLG to $10M ARR→ Browse all 60 playbooks→ Find a growth tool for live data or execution→ Get Gingiris advisory or a custom AI growth employee→