Why showing up in ChatGPT is not classic SEO
ChatGPT has no SERP, no positions to occupy, no CTR to optimize. Showing up in a ChatGPT response means your brand is cited by name in the model's synthesized text. That citation can come either from training memory (fixed corpus) or from browse / search mode (real-time web). Optimizing one does not automatically optimize the other.
Step 1 — Identify your target prompts
List 30 prompts your prospects might ask ChatGPT during a buying cycle. Sources for the language: Search Console (queries driving traffic), Reddit (industry discussions), support conversations (customer questions). 40 % discovery prompts ("best X provider"), 25 % comparisons ("A vs B"), 20 % technical, 15 % brand-explicit.
Step 2 — Build third-party authority first
ChatGPT prefers independent authoritative sources over brand sites. 32 % of cross-LLM citations route through Wikipedia, 18 % through trade press, 14 % through established press (NYT, WSJ, FT, Bloomberg). Without Wikipedia presence + 8-15 trade press mentions in the last 12 months, you have near-zero chance of being cited on open prompts.
Earned editorial PR ($2-4k/month for a specialized PR officer) is the most profitable lever. Sponsored content has near-zero LLM impact: models discount pages marked "sponsored" or "advertorial".
Step 3 — Optimize your site for extraction
Reformulate strategic H1s as questions, deploy schema.org Organization with sameAs pointing to Wikipedia, add FAQPage on product pages. Check robots.txt: GPTBot must not be blocked (verify on openai.com/gptbot that you authorize it). Add an llms.txt file at the domain root.
Step 4 — Measure and iterate
Subscribe to a GEO tool (Geoperf, Otterly, Profound). Run your panel weekly, track citation rate, source attribution, sentiment. Iterate month by month: identify prompts where you never appear, investigate why (no Wikipedia? thin press? weak structure?), fix.
Realistic timing benchmarks
ChatGPT Search (web mode): 4-12 weeks after optimization. ChatGPT standard (memory mode): 6-12 months (retraining cycle). To move fast, prioritize ChatGPT Search optimization.
Common mistakes to avoid
Blocking GPTBot in robots.txt by mimicking other bots is mistake #1 — you become voluntarily invisible. Investing only in classic SEO without touching third-party authority is mistake #2. Measuring once instead of weekly is mistake #3 — LLM stochasticity dominates on isolated snapshots.
Realistic 6-month target
For a US/UK B2B mid-market brand starting from a 5-15 % ChatGPT citation rate, reaching 30-45 % in six months is achievable with a $35-60k investment (tool + PR + on-page restructure). Above 50 % citation rate, marginal cost rises significantly and requires advanced Wikipedia strategy + quarterly flagship studies.