How to identify AI traffic in Adobe Analytics

[braulionxp]

With the implementation of AI in search engines, websites, and devices, new sources of traffic are emerging.

Currently, Adobe Analytics can identify AI traffic when it comes from "other websites" or "domains", but the rest of new AI sources is not measured by the tool, or at least not to my knowledge.

My question for the community is to know if there is something we can do with the current Adobe Analytics features to identiy these new AI sources, maybe a new configuration? or if the company is planning to launch new features to be able to calculate AI traffic?

Thanks for reading or answering my post.

[Elanchezhiyan_E]

1. Referral Domains with Custom:
You can leverage Adobe Analytics' referral domain tracking to recognize traffic from recognized AI sources such as chatbots or AI search engines.

Modify your marketing channel processing rules to label these referral domains as a separate channel, e.g., "AI Traffic."

2. Campaign Tracking Codes:

If you're advertising content via AI platforms, add UTM parameters or custom tracking codes to URLs to recognize traffic from these sources.

3. Custom Dimensions or Variables:
Create custom dimensions or variables in Adobe Analytics to record AI-specific metadata traffic. For instance:
Utilize JavaScript on your website to identify user-agent strings that are related to AI crawlers or bots and forward them into a custom variable.
Recognize any distinctive features of AI traffic (e.g., certain query parameters) and record these in Analytics.

4. IP Filtering:
Most AI platforms utilize recognizable IP ranges. You may want to set up a mapping table of the ranges and flag sessions coming from them.

5. Behavioral Analysis:
Utilize Adobe Analytics' anomaly detection or segmentation capabilities to segment out out-of-the-norm patterns of traffic, for example:
Elevated bounce rates with low session lengths.
Uncommon navigation sequences not consistent with human patterns.

Adobe could introduce features to more effectively deal with AI-generated traffic, with its increasing footprint. You can:

Engage Adobe Support/Community:
Discuss your use case with Adobe's product team or via the community forums to lobby for more advanced AI traffic identification features.
Keep yourself informed with Adobe's feature release notes to track any release announcements on AI tracking.

Machine Learning in Analytics:
Adobe's AI product, Adobe Sensei, might later provide insights on AI traffic identification. If offered, use these features for improved attribution.

What can be done?
Server-Side Tagging:
With server-side tagging, you will be able to log and analyze headers, user-agents, and other indicators that could suggest AI-generated traffic.
Use server-side logic to segment and distinguish human and AI traffic.

Cooperation with AI Platforms:
Cooperate with AI platforms such as OpenAI or search engines such as Bing Chat to understand how they display your content and whether there are any identifiers which you can utilize to monitor traffic.

Third-Party Tools:
Consider employing solutions such as bot-detection tools or analytics platforms that specialize in monitoring AI traffic, and combine these data with Adobe Analytics.


Though Adobe Analytics has yet to incorporate native support for monitoring AI traffic, the options found within the platform, as well as with innovative configurations, enable you to recognize and track such traffic. The combination of technical settings, server-side arrangements, and consistent pushing for future feature development ensures that your analytics keep pace with AI technologies​

[Sukrity_Wadhwa]

Hi @braulionxp,

Were you able to resolve this query with the help of the provided solution, or do you still need further assistance? Please let us know. If any of the answers were helpful in moving you closer to a resolution, even partially, we encourage you to mark the one that helped the most as the 'Correct Reply.'
Thank you!