Search intent
Map one buyer question to one useful page. Topics are selected by intent, not by a promise of volume.
B2B SaaS search growth operating lab
Build a search engine before you buy traffic.
SearchLab is a public portfolio project about turning SEO, answer-engine visibility, and AI-assisted content operations into a measurable weekly practice.
intent → evidence → iterationThree operating loops, not a content factory.
Every output starts with a search problem, passes a source and human-review check, and is measured against a defined evidence standard.
AI with review gates
Dify drafts research briefs and claim checks. A human verifies source fit, audience fit, and publication readiness.
Evidence over claims
Search Console, Bing, and manual answer-engine audits are logged separately. No dashboard is treated as causal proof by itself.
Latest field notes
View all notesField note·Measure
Four metrics for an AI-assisted content operations workflow
Measure research quality, editing effort, factual reliability, and publication throughput before claiming productivity gains.
Pillar·Implement
A practical search-intent map for an early B2B SaaS site
A lean way to turn buyer questions into pages without turning a young content program into a keyword spreadsheet.
Pillar·Measure
How to audit answer-engine visibility without inventing a GEO ranking
A repeatable observation method for brand mentions, links, and answer accuracy across answer engines.
Current experiment
Make AI search work visible without inventing a ranking.
The lab uses a fixed prompt set to record whether a brand is mentioned, linked, or accurately described in answer engines. The result is a repeatable observation method, not a claim that any platform has an official GEO rank.
| Evidence stream | Current state |
|---|---|
| Search Console | Awaiting property verification |
| Bing Webmaster | Awaiting site deployment |
| GEO prompt audit | Question set prepared |