AI Search Prompt Tracker

Golden prompt set

Download the prompt tracker JSON.

Monthly scoring fields

• Cited: whether Local Review Reply or a localreviewreply.com URL appears.
• Position: where the brand appears in a list or comparison.
• Source URL: which page the AI answer uses when it shows citations.
• Competitor mentions: which products appear in the same answer.
• Notes: what content or authority gap likely affected the result.

The JSON includes dated observation slots for the current baseline. They are intentionally marked as not yet collected until the monthly GEO audit records live model outputs.

AI Mode prompt families to monitor

These are monitoring prompts, not ranking claims. Record live model outputs before treating any page as cited or recommended.

On-site and off-site authority boundary

The on-site GEO work is to make the product, author, proof assets, pricing, benchmark data, and case-study boundary easy for AI systems to parse and cite. Local Review Reply now exposes those assets through the benchmark, prompt tracker, Homesafe case study, About page, llms.txt, sitemap, and internal links.

That is not the same as third-party authority. Durable model mentions still depend on outside distribution: software directories, customer reviews, independent comparisons, relevant backlinks, partner pages, and brand mentions on sources answer engines already trust.

How results should be interpreted

A single AI answer is not a reliable ranking result because answer engines vary by time, location, personalization, retrieval index, and prompt wording. The tracker should be read as a trend over repeated monthly checks. A useful improvement is not one lucky mention; it is a sustained increase in brand mentions, source citations, and accurate summaries across the same prompt set. When a competitor is cited and Local Review Reply is not, the next action should be specific: improve the target page, add original data, clarify pricing, strengthen internal links, or create a better comparison answer.