コミュニティアチーブメントバーを展開する。

Submissions are now open for the 2026 Adobe Experience Maker Awards
解決済み

Impact of AI agents & chatgpt

Avatar

Level 1

Hi,

 

Is there a way i can see what is the impact of AI/chatgpt search?


Can i see when there is an ai-agent visiting the website?

1 受け入れられたソリューション

Avatar

正解者
Community Advisor

If you're trying to measure traffic influenced by AI platforms (like ChatGPT or Perplexity), the easiest starting point is to use Referrer Domain in your reports. Look for domains like chat.openai.com, chatgpt.com, or perplexity.ai. These will show up when a user clicks a link from an AI-generated result and lands on your site.

 

If you're looking to identify actual AI crawlers or agents (i.e., automated tools accessing the site), it's trickier. Adobe doesn’t natively log user-agent strings, but if you’re capturing IPs or using a custom dimension for user-agent (via eVar or prop), you might be able to flag known AI patterns from those fields.

 

Another workaround: monitor traffic patterns where engagement is low (e.g., 1 page per visit, 0 sec time spent), and cross-reference that with unusual referrers or user-agent values. You can then tag or exclude them based on a bot-like behavior segment.

 

We haven’t seen a huge wave of traffic from these sources yet, but it’s definitely growing — worth keeping an eye on!

元の投稿で解決策を見る

2 返信

Avatar

Community Advisor

Hi @thomass59572841 I assume you want to see how many visitors are being referred by AI and/or how many hits are coming directly from AI sources themselves? You should be able to do both. 

For visitors being referred by AI, you can use the referrer domain dimension. 

For hits from AI themselves, use the domain dimension.

Here are a couple I found quickly:
chatgpt.com, perplexity.ai

Avatar

正解者
Community Advisor

If you're trying to measure traffic influenced by AI platforms (like ChatGPT or Perplexity), the easiest starting point is to use Referrer Domain in your reports. Look for domains like chat.openai.com, chatgpt.com, or perplexity.ai. These will show up when a user clicks a link from an AI-generated result and lands on your site.

 

If you're looking to identify actual AI crawlers or agents (i.e., automated tools accessing the site), it's trickier. Adobe doesn’t natively log user-agent strings, but if you’re capturing IPs or using a custom dimension for user-agent (via eVar or prop), you might be able to flag known AI patterns from those fields.

 

Another workaround: monitor traffic patterns where engagement is low (e.g., 1 page per visit, 0 sec time spent), and cross-reference that with unusual referrers or user-agent values. You can then tag or exclude them based on a bot-like behavior segment.

 

We haven’t seen a huge wave of traffic from these sources yet, but it’s definitely growing — worth keeping an eye on!