The Sivlo library
Learn the method before you argue with the number.
Everything Sivlo publishes about how it measures, what it counts, and where the evidence stops — written for the person who has to defend the result internally.
Guides and references
Depending on what you need.
The guide is the practical answer and the methodology the long one. When you would rather measure your own market than read about it, the plans come last.
- 01 / Guide
Does your business appear in ChatGPT?
A practical guide to AI visibility, GEO and measurement limits.
Read the AI visibility guide - 02 / Buyer’s guide
How to choose an AI visibility tool
Compare engine coverage, recommendations, citations, repeatability and the work behind a useful result.
Read the buyer’s guide - 03 / Product
AI visibility and recommendation tracking with Sivlo
Track business mentions, recommendations, competitors and cited sources in AI answers. Explore Sivlo’s measurement approach, coverage and limits.
Read more - 04 / Guide
SEO, GEO, AEO and LLMO: what is the difference?
Compare search engine optimization, generative engine optimization and answer visibility. Choose the outcome and measurement before the tool.
Read more - 05 / Guide
How to measure your business’s ChatGPT visibility
Define buyer questions, distinguish search and API results, inspect sources and track recommendations comparably. A practical guide for businesses.
Read more - 06 / Guide
Why does AI recommend your competitors?
Investigate buyer fit, company facts and source coverage before drawing conclusions. A practical way to diagnose missing AI recommendations.
Read more - 07 / Guide
AI visibility metrics and what they actually tell you
Mention rate, recommendation presence, citations and customers: count different outcomes separately and recognize small-sample limits.
Read more - 08 / Use case
AI visibility monitoring for marketing agencies
Build a clear client AI visibility baseline, an evidence-backed worklist and comparable reporting. Explore an agency workflow with Sivlo.
Read more - 09 / Use case
AI visibility for ecommerce and product brands
Investigate product fit, price and availability facts, and buying situations in AI answers. Define useful ecommerce visibility monitoring.
Read more - 10 / Research method
A reproducible protocol for measuring AI recommendations
A published measurement protocol covering question panels, service settings, repeats, missing observations and reporting. A method, not a claimed benchmark result.
Read more - 11 / Integration
Prepare and approve WordPress page changes with Sivlo
Explore Sivlo’s workflow for bounded WordPress page sections: connection, draft, approval, publication and public verification.
Read more - 12 / Comparison
Prompt tracking or recommendation intelligence?
Compare two workflows: monitoring a defined prompt set and investigating recommendations through buying situations. Choose for your actual needs.
Read more - 13 / Guide
AI search crawlers: ChatGPT, Claude, Perplexity and Google
Separate search crawlers, user-directed retrieval and training. Check robots.txt, indexing and firewalls while keeping private pages protected.
Read more - 14 / Reference
How the measurement works
How questions are chosen, how a recommendation is counted, what is excluded from a denominator, and how uncertainty is reported rather than rounded away.
Read the methodology - 15 / Guide
Finding your way around
From your first company review to reading a single answer, choosing an action and following what happens after you act on it.
Open the product guide - 16 / Plans
What each plan measures
Every plan carries the whole method. What changes is how much of your market is measured, how often, and how many people on your team read it.
Compare the plans
The vocabulary
Five words that carry the result.
Each one is a counting rule rather than a marketing term, and each means exactly the same thing everywhere it appears in the product.
- Recommendation Presence
- The share of eligible, recommendation-oriented answers that recommend your company. Shown with its numerator, its denominator and the number of observations behind it, so the figure can be argued with.
- Mention Rate
- The share of relevant answers where your company appears at all. A mention is not a recommendation, and the two are never added together.
- Eligible observation
- An answer that was actually returned and actually relevant. A failed provider call is an operational miss and never becomes evidence that your brand was absent.
- Comparable measurement
- Two runs sharing a market-model version, a buying-situation set, a sampling strategy and a provider configuration. When any of those change, Sivlo says the runs are not comparable rather than drawing a line between them.
Still have a question
Ask it, and we will answer with the working.
If something in the method does not hold up against your market, we would rather hear it than have you wonder. Tell us which page you were reading and what did not add up.