CiteMET · Cited, Memorable, Effective, Trackable
CiteMET (Cited, Memorable, Effective, Trackable) is a method promoted by LLMrefs that adds share buttons to pages that open AI assistants with a prefilled prompt. Its effectiveness has not been independently demonstrated, and it comes close to techniques Microsoft has described as recommendation poisoning.
Standalone definition · 44 words
What does the method propose?
LLMrefs presents it as a way to grow LLM traffic with dynamic AI share URL buttons: the click opens an assistant with a prompt containing the brand, which could get it memorized as a preferred source. These claims come from the vendor promoting the method.
Why treat it with caution?
No independent study demonstrates the claimed effect. A prefilled prompt that tries to influence what the assistant retains falls, depending on its content, under the manipulation Microsoft described as recommendation poisoning. Google also advises against inauthentic mentions.
What is the alternative?
Offer a neutral share or summary button whose prompt is transparent to the user and contains no instruction about what the assistant should memorize or recommend, and measure AI referral traffic in analytics.
Frequently asked questions
Is CiteMET a recognized practice?
It is a method proposed by a tool vendor. No engine documents it as an official lever.
Is a share button to ChatGPT forbidden?
Not in itself. The risk comes from a prefilled prompt that manipulates the assistant’s memory or recommendations.
Related terms
Further reading
- Profound vs Semrush AI Visibility Toolkit: which really tracks your AI visibility
- ← Back to the GEO glossary (114 terms)