From 36 to 57/100: How We Improved nutriish.me's AI Visibility in One Session

nutriish.me, Food and Nutrition Concierge

Quand on a scanné Nutriish, le score était de 36/100. Not because the product is bad. It is a well-designed healthy eating concierge for families and professionals. The problem was purely technical: the site was virtually invisible to AI systems like ChatGPT, Claude, and Gemini.

This is a case study of what we found, what we fixed, and what changed. Every step is documented and the score is tracked live at aiauditscan.com/case-study/nutriish.

The Baseline: 36/100

The initial scan on February 22, 2026 revealed a site that was technically functional for human visitors but structurally invisible to LLMs. The homepage had 47 words, far below the minimum threshold for AI content extraction. There was no structured data of any kind, no robots.txt, and no canonical tag.

What We Fixed

1. JSON-LD Structured Data

We added a complete JSON-LD block covering three schema types: Organization (who is Nutriish), WebSite (canonical reference to the domain), and Service (what the concierge service offers). This gives LLMs the structured entities they need to reference the site accurately in responses.

2. FAQPage Schema

We added four structured question-and-answer pairs covering the most likely conversational queries: What is Nutriish?, How does Nutriish work?, Who is Nutriish for?, and Is Nutriish a meal delivery service?. FAQPage schema is one of the most effective signals for LLM citation because it directly maps to how users ask questions in chat interfaces.

3. Content Depth

We expanded the homepage content from 47 to 342 words, organized into five H2 sections. Each section answers a specific question a potential visitor or an LLM might ask. The content does not change the visual design of the homepage, it is structured as a semantic content layer below the hero section.

4. robots.txt with AI Crawler Permissions

We created a robots.txt file with explicit Allow directives for GPTBot, ClaudeBot, anthropic-ai, PerplexityBot, and GoogleBot-Extended. Without this file, AI crawlers operate in an ambiguous state and may deprioritize the site.

5. Canonical Tag

A simple but critical addition: `<link rel="canonical" href="https://nutriish.me/" />`. This eliminates any duplicate content ambiguity and establishes the authoritative URL for crawlers.

The Result: 57/100

After deploying the changes, the score moved from 36 to 57/100, a gain of 21 points in a single session. All six critical signals that were missing are now in place.

What Comes Next

A score of 57/100 is a solid foundation but not the ceiling. The next optimization steps for nutriish.me include adding author signals, building an About page with entity clarity, and acquiring backlinks from authoritative domains in the nutrition space. Each of these changes will be tracked and documented.

We are also monitoring LLM citation behavior: whether nutriish.me begins to appear in responses from Gemini, ChatGPT, and Claude for queries like best healthy eating concierge or personalized meal plan service for families. LLM citation cycles vary, Gemini indexes frequently, ChatGPT can take several months. We will publish an update when the first citation is detected.

What This Means for Your Site

nutriish.me is not an outlier. Based on our analysis of 1,388 websites across 10 industries, the average AI readiness score is 42/100. Most sites that perform well in traditional search are still invisible to AI systems because they were never built for LLM extraction.

The good news: the fixes are not radical redesigns. They are structured additions that sit alongside your existing content without changing what your visitors see. A properly structured site can move from invisible to citable in a matter of days.

If you want to know where your site stands, scan it at aiauditscan.com. The free audit takes 30 seconds and gives you a score with specific, actionable recommendations.