The author observed 145 bots visiting their site for two weeks, creating a scoring system to analyze each bot's behavior. They found that most traffic was not human, and AI crawlers (like GPTBot and ClaudeB) and major search engines were among the best-behaved.
The author implemented a tracker to monitor which AI bots access their new static website, analyzing CloudFront logs stored in S3. After five days, bots like OAI-SearchBot and Bingbot were identified, with the methodology based on user-agents.
Este conteúdo aborda a detecção da intenção do usuário em websites através da análise de sinais comportamentais, que é o fundamento da IA comportamental. Ele explica como a análise de padrões de interação, como movimento do mouse e rolagem, permite inferir a intenção do visitante com alta precisão.
The author conducted an experiment using seven different AI-citation trackers to determine their website's visibility to AI search. The trackers yielded vastly different citation counts, with the author highlighting the inconsistency and preferring a tool for its transparency rather than its accuracy.