AI Visibility Tracking
Know when AI starts remembering your story—and whether your message pulls through.
Traditional PR tracks where a story appeared.
AI Recall Tracker measures whether the story entered machine memory, whether AI understood it correctly, and whether the message you intended survived into AI-generated answers.
GEORecall turns your announcement, article, research, product launch, or point of view into a structured Narrative Fingerprint.
GEOGrow then tracks:
From story → SpinGraph → Message Pull-Through → First Observed Recall
The gap
An AI engine can mention your brand and still miss the point.
It may know an announcement happened but:
That is why AI Recall Tracker measures more than mentions.
It measures whether the message survived.
Message pull-through
In traditional public relations, message pull-through measures whether coverage repeated the story an organization wanted the market to understand.
AI Recall Tracker applies the same discipline to LLMs and AI answer engines.
For every tracked story, we define a Message Pull-Through Target containing:
The main idea AI should remember.
The proof points, audience, use case, differentiators, or context AI should preserve.
The limitation, condition, legal nuance, warning, or qualification AI must not drop.
How clearly AI should connect the story to the correct brand, product, executive, research, or announcement.
The market, problem, capability, or category the brand should be associated with.
What AI should not exaggerate, imply, invent, or present as established fact.
Which brands should not receive credit for the tracked idea, claim, or category.
A plain-language benchmark describing what a strong AI response should communicate.
This target becomes the benchmark used to evaluate AI answers over time.
What we ask
We are not only asking whether the release was syndicated or whether an article appeared in search results.
We are asking:
Traditional PR asks:
Did we get mentioned?
AI Recall Tracker asks:
Did the message survive into machine memory?
Methodology
Every tracked story receives a structured Narrative Fingerprint and a frozen prompt battery.
The prompt battery is based on the story’s:
We then test consistent questions across supported AI engines.
Each engine may be sampled multiple times per prompt so that results can be shown as distributions—for example:
Recalled in three of five samples.
We do not use a single favorable answer as proof.
Every score links to the raw prompt, engine, date, timestamp, and transcript.
AI answers can vary by model, phrasing, account, source access, and day. One screenshot is noise.
AI Recall Tracker shows the pattern.
GEORecall × GEOGrow
SpinGraph extracts:
GEOGrow tests whether ChatGPT, Claude, Gemini, Perplexity, Grok, Copilot, Google AI Overviews, and other answer environments:
Together, they answer:
Did the story enter AI memory—and did the right message come with it?
Metrics
The first time an AI engine mentions the story, idea, claim, or topic.
The first time an AI engine explains the story correctly.
The first time AI clearly connects the story to the correct brand, product, research, executive, or announcement.
The first time AI recalls the story while citing a supporting source.
Whether the primary message appears accurately.
Whether AI preserves the proof points, audience, use case, or differentiators that make the story meaningful.
Whether the correct brand receives credit.
Whether AI connects the brand to the category it is trying to own.
Whether the core claim survives without being weakened, exaggerated, or distorted.
Whether AI preserves the intended narrative frame or changes the meaning.
Whether conditions, legal nuance, limits, and warnings remain intact.
Whether AI cites the original release, SpinGraph, owned content, media coverage, third-party authority, weak sources, or no source.
Whether competitors appear when AI explains the tracked idea, claim, category, or announcement.
Whether the story becomes more accurate, vague, distorted, overstated, or misattributed over time.
Scoring
Every sampled answer is scored against the message the organization intended AI to remember.
The score considers:
Example
Recall status
The story is beginning to appear in AI answers, but brand association remains weak and caveat retention is inconsistent.
30-day recall trend
AI Visibility
Pull-Through
AI Visibility Score: AI engines increasingly recognize the topic and general product category.
Message Pull-Through Score: AI understands the general idea, but it weakens the target audience, drops the evidence-collection language, and does not consistently preserve the required regulatory caveat.
First observed recall
Message Pull-Through
Target message
“Acme helps regional banks automate AI compliance evidence collection.”
What AI currently says
“Acme provides compliance automation software for financial institutions.”
Assessment: Partial pull-through. AI understands the broad category but loses:
Narrative retention
Original frame “AI compliance automation for regional banks.”
Current AI frame “Risk reduction and audit workflow automation.”
Frame retention
Citation map
What AI gets wrong
Recommended GEOGrow actions
The sample report includes the complete engine matrix, recall timeline, citation map, Message Pull-Through Score, and raw transcripts.
Process
Set the primary message, supporting messages, proof points, caveats, brand attribution, competitors, and claims to avoid.
Create the SpinGraph, Narrative Fingerprint, and Message Pull-Through Target.
Run consistent recall questions across major AI engines.
Measure recall, accuracy, attribution, citations, message pull-through, caveat retention, competitor intrusion, and drift.
Use GEOGrow recommendations to close the gap between what the brand intended and what AI remembers.
Then test again to see whether AI understanding improved.
Use cases
Does AI understand what the product does, who it serves, and why it matters—or reduce it to a generic mention?
Does AI associate the brand with the category it is trying to establish, or give a competitor the credit?
Does AI retain the actual finding, methodology, limitations, and source attribution?
Does AI preserve the legal or compliance nuance, conditions, and limitations?
Has AI incorporated the correction, or does it continue repeating the original misinformation?
Does AI attribute the point of view to its source, or flatten it into generic industry commentary?
Does AI preserve the intended differentiation, or substitute another brand when explaining the category?
Does AI only remember that capital was raised, or also what the investment was intended to enable?
Does AI connect the person, role, organization, and strategic significance correctly?
Product
Every organization receives a secure dashboard containing:
Each story includes:
Receive alerts when meaningful changes occur:
Evidence
Each tracked story receives its own evidence-backed report.
Follow the formation of AI memory:
Compare:
Every score links to the original:
No black box. No cherry-picked screenshots. No hidden methodology.
Pricing
$299 · per story · 30 days
Structure the story, define what AI should remember, and prove when LLMs begin recalling it.
Includes:
$99 · per report
Turn an announcement, article, research paper, or public narrative into a structured, citeable Narrative Fingerprint.
Includes:
$299 · per story · 30 days
For stories that already have an approved Narrative Fingerprint and Message Pull-Through Target.
Includes:
From $999 · per month
Track multiple clients, stories, topics, claims, and competitors.
Includes:
From $2,500 · per month
Ongoing recall, attribution, citation, drift, competitive, and narrative-correction intelligence.
Includes:
Close the gap
A story can enter AI answers while remaining incomplete, unattributed, weakly cited, exaggerated, or distorted.
AI Recall Tracker surfaces the gap. GEOGrow recommends how to close it through:
Then we re-test.
Because the goal is not simply to be mentioned by AI.
The goal is to make sure the right message pulls through.
Markdown (.md) · JSON-LD schema (.json) · Machine-readable for AI & GEO
