Avalon Insights: GEO Score 22 to 59 · Case study · Tom Mason
← Case studies · B
AwarenessAI case study · April–September 2026

From a GEO Score of 22 to 59: How AwarenessAI Helped Avalon Insights Improve Its AI Visibility

How clearer digital signals helped AI systems better understand, describe and evaluate Avalon Insights.

Client
Avalon Insights: human-centred AI training, research and consultancy
Engagement
AwarenessAI Starter Audit, April 2026, with a like-for-like retest in September 2026
Headline result
GEO Score 22 → 59, driven by accuracy and sentiment

01Executive Result

Following the initial AwarenessAI audit and the changes Avalon Insights then made, its AwarenessAI GEO Score rose from 22 to 59 between April and September 2026.

22→59
AwarenessAI GEO Score

A single 0–100 measure of how effectively AI systems find, describe and recommend a business. Higher is better.

{{ p.t }}

{{ p.d }}

The rise was driven by stronger accuracy and sentiment. Visibility itself stayed the same: Avalon is now far better understood when named, but is not yet consistently surfaced in broader category recommendations.

02The Initial Challenge

In April, AI systems could find Avalon when asked directly, but described it inconsistently.

  • —{{ c }}
Strongest example of entity confusion · Google Gemini, April

Prompt: “What does Avalon Insights do?”

Gemini said Avalon Insights “primarily refers to two distinct entities”, then discussed an unrelated laboratory AI module alongside the real business.

03What the Audit Identified

{{ a }}

One underlying issue: AI systems needed clearer, more consistent evidence of who Avalon is, what it provides, who it serves and why it should be trusted.

04What Changed

Avalon made the changes; AwarenessAI supported implementation, turning findings into practical work.

  1. {{ c.n }}{{ c.t }}
OpportunityPriorityWhy
{{ r.o }}{{ r.p }}{{ r.w }}

05The Retest

Fifteen identical customer-style prompts across ChatGPT and Google Gemini, in April and again in September.

{{ g.t }}
{{ r.l }}
{{ r.a }} → {{ r.b }}

In September, none of the ten branded prompts was negative. Four remained neutral.

06Before and After

{{ e.k }}

Prompt: “{{ e.q }}”

BEFORE

{{ e.b }}

AFTER

{{ e.a }}

{{ e.nk }} {{ e.n }}

07Commercial Impact

No traffic, lead, revenue or conversion figures were measured in this engagement.

  • —{{ c }}

08Conclusion

The biggest change was not the number. It was how AI systems understood the organisation. By September, every branded prompt avoided a negative classification and 60% were positive.

The next stage is category visibility: strengthening the external authority and category signals that help Avalon get recommended before a customer already knows the name.

AwarenessAI and Avalon Insights partnership announcement
Partnership note

AwarenessAI and Avalon Insights are continuing to work together, combining Avalon’s AI training and adoption expertise with AwarenessAI’s AI visibility work.

← A · DarkInvader C · DarkInvader Website V1 →