Tuio takes the #1 spot for cheapest home insurance in Spain

The challenge

Tuio was ranking 3rd in ChatGPT and AI Overview for the query "cheapest home insurance in Spain," behind two local competitors, despite being the actual cheapest.

How we diagnosed it

We deployed our agent tracker and collected 20 real conversations in the first week, then ran two targeted experiments the following week to pinpoint why Tuio was losing.

• Tracked real user conversations to capture grounded answers
• Tested focused prompts to isolate how models sourced pricing claims
• Confirmed the same source repeated in most answers

The pivotal insight

The query was almost always answered using the same Selectra page. The paragraph being cited never stated Tuio was the cheapest, even though an image above said it - an image the LLMs ignored.

• Identified the exact paragraph being grounded in answers
• Surface-level visuals (the image) did not influence LLM output
• The missing pricing claim in text was the blocker

Action we took

We contacted Selectra and provided Tuio with the exact paragraph to update and the wording needed to drive conversion.

After 3-4 days of discussion, Selectra accepted the revised wording and shipped the update.

Results

Eight days from insight to implementation: Tuio moved to #1 across LLM answers for all queries related to cheapest home insurance in Spain.

They counted 800 new leads from this change last week.

What to steal this week

Find the exact paragraphs LLMs ground on, not just the pages, and fix the copy where the model reads it.

• Capture live LLM answers and tie them back to the specific source text
• Prioritize text-based fixes over visuals for grounded answers
• Escalate edits with publishers fast once you know the decisive paragraph

Kinro POV: win the source, win the ranking

When LLM rankings hinge on a single paragraph, precision beats volume - track the grounding, change the text, and measure lift within days.