Contact by email
Back to Blog
AI News

An AI-visibility baseline a board can read

GlasBox Studio · · 3 min read

TL;DR: Key Takeaways

Measurement without a frame produces slides. A frame has four columns: question (commercial, not editorial), system (organic Google, AI Overview, ChatGPT, Gemini, Perplexity), finding (named / sourced / wrong), priority (revenue, risk, effort).

  • The baseline needs a cap.
  • Errors matter more than hits.
  • SOURCE/01 by GlasBox (Full) delivers that matrix plus a 90-day order.

Measurement without a frame produces slides. A frame has four columns: question (commercial, not editorial), system (organic Google, AI Overview, ChatGPT, Gemini, Perplexity), finding (named / sourced / wrong), priority (revenue, risk, effort).

The baseline needs a cap. Ten to fifteen questions per language is enough to start. A hundred prompts with no decision is theatre.

Errors matter more than hits. If the model puts you in the wrong city, service or price band, that is a knowledge problem - often on your own site or in stale directories, not "in the algorithm".

SOURCE/01 by GlasBox (Full) delivers that matrix plus a 90-day order. The free short check does less on purpose: one domain, three questions, three observations. Mixing the two misleads the buyer.

---

Measuring ChatGPT visibility
Prompt set for monitoring
Perplexity and visible sources

Next step

GlasBox measures classic SEO and AI visibility separately. No ranking or mention guarantee. The free short check stays thin on purpose.

Request an introductory call

Frequently Asked Questions

What belongs in a baseline a board can read?
Four columns: commercial question, system, finding, priority by revenue, risk and effort.
How many questions are enough to start?
Ten to fifteen questions per language. A hundred prompts with no decision is theatre.
What matters more than hits?
Errors. The wrong city, service or price band is often a knowledge problem on your own site or in directories.