AI visibility tracking keeps chasing a moving target, because the AI systems people ask for recommendations keep changing shape. The newest shape is Meta’s Muse, a personal AI agent rather than a chat window, and a visibility platform adding coverage for it within two weeks says as much about where marketing measurement is headed as the feature itself does.
What happened
Peec AI announced on September 23 that it now tracks brand visibility inside Muse Spark, the large language model powering Meta’s Muse personal AI agent, which launched September 8. The addition lets marketing teams measure brand visibility, sentiment and cited sources inside Muse alongside Peec AI’s existing coverage of ChatGPT, Google AI Mode, Gemini, Perplexity, Copilot, Claude and several other answer engines, all from one dashboard. “Tracking the Meta Spark LLM in Peec AI is the closest approximation brands have to estimate their performance in Muse, Meta’s viral AI assistant,” said Malte Landwehr, Peec AI’s chief product and marketing officer.
Why it matters
Every prior wave of AI search visibility tooling has tracked systems that describe options: a chatbot answers a question and cites sources. Muse is built to act on a user’s behalf, including tasks like purchasing and negotiation, so the signals shaping its output influence what gets chosen, not just what gets described. A brand invisible to an answer engine loses a citation. A brand invisible to an agent that acts loses the transaction.
The original insight
The speed of the response, two weeks from Muse’s launch to dedicated coverage, is itself the story: AI visibility vendors now race to instrument each new agent the moment it ships rather than wait for one to prove durable, because no single assistant’s share of AI prompts can be assumed to hold. That argues against a CMO picking a visibility tool built around today’s dominant assistant, and for one whose roadmap already chases the next platform first. The category this site has called martech’s core discipline now has to cover agents that transact, not just assistants that answer.
Source: Peec AI
