Notes from the picks-and-shovels side of AI search.

ChatGPT searches the web before it answers, but cites only 15% of the pages it reads. How retrieval and citation work, what gets cited, and how to show up.

No, and the reason matters for your brand. Why ChatGPT answers vary by user, settings, and time, even at temperature 0, and why one screenshot tells you nothing.

Answer engine optimization vs generative engine optimization: how they differ, where they overlap, whether they are the same thing, and which to prioritize.

SEO got you found; AEO gets you quoted. What answer engine optimization is, how it works, AEO vs SEO, AEO vs GEO, and how to start getting cited in AI answers.

The GEO market split into tools, agencies, and APIs. A comparison of the best GEO platforms, agencies, and the API option, by what they do and who they fit.

Search used to mean a spot in the ten blue links. Now it means being one of the few sources an AI cites. What GEO is, how it works, GEO vs SEO, and how to start.

See how ChatGPT, Perplexity, and Gemini describe your brand. A free, step-by-step AI visibility audit you can run in 20 minutes, plus the best free checker tools.

You can rank #1 on Google and still be invisible in ChatGPT. What AI visibility means, how it differs from SEO, and how to measure it across every engine.

Profound, Peec, Otterly, Ahrefs, Semrush, and the API option compared by coverage, price, and who they fit. A dozen tools, $300M+ in funding, prices from $29 to $499/mo.

The AI answer box now sits above the blue links on a huge share of searches. What Google AI Overviews are, how they work, what they do to traffic, and why being cited matters.

Ranking #1 no longer wins the answer. The 2026 playbook for getting cited in Google AI Overviews: self-contained passages, schema, entity authority, and how to track your citations.

ChatGPT, Perplexity, Google AI Mode, Gemini, Copilot, and Grok ranked by accuracy, citations, and use case. Only 11% of cited sources overlap between engines, so the right pick depends on the job.

The complete build of a Generative Engine Optimization tool. Four atoms, six surfaces, daily cadence, real cost numbers. One weekend of work, ~$120 a month to run.

Track which URLs AI engines cite as answers. Citation extraction, gap analysis, and an outreach feedback loop. Around 150 lines and ~$50 a month.

Build the single visibility score every CMO asks for. One number, plotted over time, sourced from real AI answers across six surfaces. ~80 lines, $30 a month.

Track brand position inside ChatGPT, Claude, Gemini, and Perplexity answers the same way you track Google rankings. ~100 lines, $15 a month.

Compare how ChatGPT, Claude, Gemini, and Perplexity treat your brand differently. Per-model gap reports drive specific optimizations the rollup score hides.

Build an AEO and GEO automation agent on Hermes in one evening. Five steps, real config, $15 to $30 per month to run.

Six concrete automations you can run on Hermes today: visibility checks, citation gap analysis, share-of-voice tracking, AI Overview alarms, prompt discovery, and content evals.

Wire MentionsAPI into Hermes Agent and watch every ChatGPT, Claude, Gemini, and Perplexity answer about your brand. Setup in 10 minutes, from $0.02 per check.

A 1,000-query teardown of how badly the LLM APIs you are optimizing for diverge from what real users actually see in the UI. With numbers, sources, and the picks-and-shovels framing.