Evaluate an optimization claim by its evidence, scope, and proposed mechanism. The examples below distinguish documented behavior from conclusions the evidence does not support.
"SEO is dead"
SEO remains relevant where an AI product retrieves information through search. Google explicitly applies existing SEO requirements to its AI search features. A search position does not guarantee a generated recommendation. See SEO and AI Search Optimization.
"Blocking GPTBot protects your content"
GPTBot is a specific crawler with a documented purpose. A rule for that agent does not control every provider, remove previously learned information, or prevent discussion on third-party sites. Decide access according to the intended use and the relevant provider documentation.OpenAI's bot documentation, which states each bot's robots.txt setting is independent of the others.
"Schema makes AI cite you"
Structured data can support documented search features and clarify page information. It does not guarantee citation. Check the specific product’s requirements and keep markup accurate. See Structured Data and Machine-Readable Facts.Google's structured data introduction, which describes structured data as clues about page meaning that can enable rich results; it makes no claims about AI citation.
"llms.txt is a proven standard"
llms.txt is a proposal. A product may support a proposal for a particular use without making it a universal search requirement. Ask which provider documents the claimed use and what evidence establishes a visibility benefit.The llms.txt proposal, authored by Jeremy Howard, published September 3, 2024; it describes itself as a proposal.
"AI search rewards fresh content everywhere"
Update content when facts change or the explanation needs correction. A new timestamp alone does not establish usefulness. A model can also use older information from parameters or retrieved sources. Keeping Information Current covers the checks.
"You can rank #1 in ChatGPT"
A brand can appear first in an individual ChatGPT answer. That observation does not establish a stable first position across prompts, accounts, or repeated runs. Report frequency and position with the test conditions and sample size.
"Video is a major citation driver"
The value of video depends on the audience, question, and product. A low citation share in one study does not prove that video cannot help. A correlation with brand recommendations does not prove that producing video causes them.
"Mentioning your brand more on your own site teaches the model"
Repeating a brand name does not directly update a deployed model. Use the name where it makes the subject clear. Do not publish repetitive, low-value pages to manipulate search or generated answers.Google's spam policies, which define scaled content abuse as generating many pages primarily to manipulate rankings rather than help users.
How to evaluate another claim
Ask what is being changed, which documented operation it could affect, and how the effect will be measured. Distinguish a provider statement, an observation, and an experimental result.