博客

Use contribution margin, attributable customers and explicit assumptions to budget AI visibility work. Citations and calculator outputs are not revenue.

Understand what llms.txt can document, what crawler logs cannot prove, and how to test a low-cost change without claiming a ranking benefit.

Check Perplexity access, relevant sources and page evidence, then measure fixed questions. A citation is observable; inclusion is never guaranteed.

Diagnose access, relevance and evidence gaps before changing content. A practical checklist that separates failed checks from measured non-citations.

以 trirankai.com 为例，记录 Search Console 导入、AI 问法配置、抓取规则修正及后续效果复查口径。

Google has said it knows about low-quality 'best of' listicles and works to combat that abuse in Search and Gemini. What it actually said, what a weak list looks like, and how to write one that holds up.