The AI Search Playbook · 2026 Edition · v1.3

40 levers to make AI search cite your brand first.

Every lever is effort-rated, cost-rated, and mapped to its prerequisites. Pick your industry and company size. The 40 levers re-rank for your context. Every claim links to verifiable primary research, with deeper reading on lawrencehitches.com for each lever.

40 levers · Content · Technical · Earned 4 execution layers · dependency-mapped Verified sources only v1.3 · Apr 2026 · changelog
88.46%
of ChatGPT citations come from the general web search index
89.78%
citation rate for natural-language URL slugs vs 81.11% for opaque URLs
500
days median age of cited pages, established beats fresh inside a retrieval set
1.93%
Reddit's citation rate despite 16M retrieval points, context layer, not citation layer
The full playbook
A stat hero showing 40 AI search levers, re-ranked for your context, with supporting points on dependencies, execution templates and cited research, underlined in lime.
40 AI search levers, re-ranked and sequenced into a living, prioritised system.

40 levers, re-ranked for your context.

Pick your industry and company size. The 40 levers re-rank with priority badges and dim the ones that don't fit. Click any lever for the full scorecard, effort, cost, prerequisites, KPI, expected timeline, and deeper reading.

Personalise the playbook

Your selections never leave your browser.
Showing all 40 levers, default priority.
Execution sequence

Which levers unlock which.

Ship in this order. Layer 0 levers have no prerequisites, start here. Each subsequent layer depends on levers completed in prior layers. Most teams try to ship Schema Slam (Lever 15) or Newsroom (Lever 40) first and get nothing from it because Layer 0 isn't done yet. Click any lever to see its full scorecard.

Changelog

This is a living document.

Every time a new primary source drops, the affected lever updates and an entry appears here. Subscribe below to get each update in your inbox.

Apr 2026v1.3

All 40 levers now have a verifiable external source

Eight levers that previously shipped with evidence-only statements (no external citation) now link to specific studies: AI conversion rate research (12M-visit study), FAQ schema → AI Overviews correlation (Frase), Google Dec 2025 Core Update E-E-A-T analysis, Core Web Vitals ranking correlation, Search Engine Land hreflang research, Zyppy 23M-internal-links study, RankPay rich snippets CTR data, Search Engine Land brand-to-links ratio research.

Apr 2026v1.2

Renamed plays → levers, stripped person attribution, added internal reading per lever

  • "Plays" renamed to "Levers" throughout
  • Person-name attribution removed; sources cite the actual content
  • Each lever now links to deeper reading on lawrencehitches.com
  • Levers without verifiable external research kept the evidence statement; source line was omitted (closed in v1.3)
Apr 2026v1.1

Industry filter, scorecards, dependency graph, downloadable CSVs

  • Every lever has effort, cost, prerequisite, KPI, and timeline ratings
  • Industry selector re-ranks the 40 levers for SaaS / ecommerce / local / B2B / media
  • Company size filter dims levers below a certain revenue band
  • "Execution sequence" section groups levers into four dependency layers
  • Three downloadable CSV templates, ecommerce, SaaS, local
Apr 2026v1.0

Initial 40-lever launch

Playbook published with 40 levers across Content, Technical, and Earned categories.

Need this executed?

I run this playbook for a small number of clients each quarter.

If you'd rather have someone implement Layers 0, 1 on your site than read 40 levers, that's what I do.

Book a 20-minute AI visibility call →
Verified sources

Every cited stat and study links to original research.

Where a lever's evidence statement could not be traced to a specific verifiable URL, the source line was omitted rather than guessed at. The list below is what's actively cited in the playbook.

Why ChatGPT Cites Pages, 1.4M-prompt study (Apr 2026)ahrefs.com/blog/why-chatgpt-cites-pages/
Google Search Central Live Toronto slides (Apr 2026)jcchouinard.com/...-toronto-slides-april-2026
Reddit B2B SaaS dominance studyfoundationinc.co/lab/reddit-b2b-saas
How Inlinks helps build topical presence on LLMsinlinks.com/insight/...-on-llms
17 Content Types That Survive Google Zerosignal.zyppy.com/p/content-google-zero
Zyppy "Winning Websites" SEO studieszyppy.com/seo-studies/
The AI Shopping Playbooksignal.zyppy.com/p/ai-shopping
Search happens everywhere, SparkToro researchsparktoro.com/blog/...-search-happens-everywhere
How ChatGPT works, semantic scoring breakdownqueryburst.com/blog/how-chatgpt-works/
What 34 studies reveal about AI searchjyscauri.substack.com/...-34-studies
Google's "creating helpful, reliable, people-first content"developers.google.com/...creating-helpful-content
AI Search Traffic Converts 5x Better, 12M-visit studysuperprompt.com/...5x-higher-than-google
Frase: Are FAQ Schemas Important for AI Search, GEO & AEO?frase.io/blog/faq-schema-ai-search-geo-aeo
Google December 2025 Core Update, E-E-A-T analysisthatware.co/google-december-2025-core-update/
Core Web Vitals & SEO, ranking correlationcorgenx.com/blog/core-web-vitals-impact-search-rankings
Search Engine Land: Hreflang, the hidden crisis in international SEOsearchengineland.com/guide/what-is-hreflang
Zyppy: 23 Million Internal Links SEO Case Studyzyppy.com/seo/seo-study/
RankPay: How to Increase Google SERP CTR with Rich Snippetsrankpay.com/rich-snippets/
Search Engine Land: The Brand-to-Links Ratio Revolutionsearchengineland.com/guide/brand-to-links-ratio-revolution