Buyer question
Can the report explain why the client was missed, misclassified, under-cited, or beaten by a named competitor?
Find the gap between AI memory and market proof.
These public audit examples show the chain agencies need before selling AI visibility work: Presence, Preference, Perception, AI Readiness, and Share of Model are shown in the same order as the product, with each audit exposing AI Memory, Web Evidence, and the gap between both.
Public audit result previewEach audit: AI Memory · Web Evidence · Gap
REFRESHEDPremium outdoor drinkware, coolers, and gear
YETI
Positive · US outdoor drinkware buyer
Generative Brand Index
47/100
Low confidence
YETI remains well understood as a premium durability brand, but the report shows a sharp proof gap: AI can remember the brand, yet unbranded Presence is weaker when the buyer asks value-sensitive cooler and drinkware questions before naming YETI.
Compared with Hydro Flask, RTIC Outdoors, Stanley 1913. 7-provider AI Memory and Web Evidence Search completed.
Complete evidence set available. Missions turn the memory/proof gap into priced agency work.
Use this as the buying filter
Before you buy another AI visibility tool, inspect whether the output can support a client decision.
Proof artifact
Does the report show Presence, Preference, Perception, AI Readiness, Share of Model, GBI context, missions, and retest targets without hiding missing evidence?
Next step
If the sample matches your sales motion, request the 5-credit kit or compare pricing before scaling to recurring delivery.
Presence
AI Memory vs Web Evidence
AI Memory
47 /100
Web Evidence
43 /100
Gap
-4 pts
Web evidence under-defends the answer
The brand is less consistently surfaced in unbranded buyer prompts; web evidence does not yet close the discovery gap.
Preference
AI Memory vs Web Evidence
Direct win rate: 71%
AI Memory
80 /100
Web Evidence
86 /100
Gap
+6 pts
Web evidence lifts the answer
Grounded evidence improves YETI's competitive recommendation strength when proof is retrievable.
Perception
AI Memory vs Web Evidence
AI Memory
71 /100
Web Evidence
72 /100
Gap
+1 pts
Web evidence lifts the answer
The named-brand story is strong in both model memory and grounded evidence; the remaining gap is proof depth for premium value.
AI Readiness
AI Memory vs Web Evidence
AI Memory
50 /100
Web Evidence
50 /100
Gap
0 pts
Memory and evidence are aligned
Support evidence is available, but schema answerability still limits how cleanly answer engines can reuse the proof.
Share of Model
AI Memory vs Web Evidence
AI Memory
64 /100
Web Evidence
65 /100
Gap
+1 pts
Web evidence lifts the answer
Derived from Presence occurrence plus Preference wins: YETI owns roughly two-thirds of measured answer space, with Web Evidence slightly improving the share.
GBI and remaining evidence gaps
What the homepage preview compresses into one result card
Generative Brand Index
47 /100
Low confidence · 100% evidence coverage
Per-audit gap
Perception is nearly balanced, Preference improves with Web Evidence, and Presence still weakens once buyers ask unbranded value-led questions.
Commercial gap
The premium value story needs more extractable durability, warranty, insulation, and cost-per-use proof.
GBI gap
GBI is 47 because Presence still drags the index even though Share of Model is calculable from occurrence and Preference evidence.
Missing evidence:
Audit structure
What the complete client audit tests
Presence
Unbranded buyer prompts show whether the brand appears before the buyer names it.
Preference
Competitive prompts measure who gets picked, against which named alternatives, and why.
Perception
Named-brand prompts show how accurately AI describes, trusts, prices, and recommends the brand.
AI Readiness
Support evidence checks whether pages, schema, facts, and source blocks are reusable by answer engines.
Share of Model
Provider-level share evidence explains how often the brand earns answer space across the measured prompt set.
GBI + Missions
The Generative Brand Index rolls the evidence into an executive score and turns gaps into retestable missions.
Mission Control
The work that changes the next answer
Done
Refresh AI Memory and Web Evidence
Perception, Preference, and Presence were rerun for the US outdoor drinkware buyer target.
Active
Strengthen price/value proof
Publish extractable durability, warranty, insulation, and cost-per-use proof blocks.
Active
Defend unbranded Presence
Create comparison-safe source blocks for cooler, tumbler, and bottle buyer prompts where competitors intercept the answer.
Queued
Retest same buyer context
Rerun the same target after source updates and compare Preference + Presence movement.
The score is only the visible edge. The valuable part is the ordered chain underneath: Presence, Preference, Perception, AI Readiness, Share of Model, GBI, missions, and the same-scope retest that proves whether the fix moved anything.
Retest plan
Retest the US outdoor drinkware buyer target after publishing price/value and durability proof blocks; watch whether grounded Preference stays above AI Memory and Presence closes the gap.
Client-ready artifact
A sample report should make the next paid action obvious.
The client does not need another vague AI score. They need to see the hidden pattern, the work your team will ship, and the exact retest target that proves movement next month.
What the client sees
- Executive summary of the current AI-answer risk
- Audit order matched to the product: Presence, Preference, Perception, AI Readiness, Share of Model
- For each audit: AI Memory score, Web Evidence score, and the gap between both
- Plain-English explanation of what the web can prove and what AI still misremembers
- Competitor preference patterns and direct-win context
- Prioritized Mission Control actions
- Same-scope retest target for the next client report
What the agency sells next
- Remediation sprint scoped from the highest-impact gaps
- Citation and source repair for facts the models cannot verify
- Content, entity, source, and AI Readiness cleanup
- Monthly same-scope retest and executive reporting retainer
- Client explanation pack that separates AI memory from current market proof
- QBR evidence pack that shows what moved, what remains unsupported, and what still blocks preference
What this public sample is
- A public preview built from named, public-market brands
- A reduced view of the client-ready artifact, not the full private evidence dataset
- A proof format agencies can use without inventing testimonials or endorsements