Expand my Community achievements bar.

SOLVED

same session has multiple user-agents

Avatar

Level 1

The session is combination of post_visid_high, post_visid_low, visit_num and hit_time_gmt. we are generating session using these 4 attributes. But we see there's been multiple user-agents associated with the same session generated with above attributes. 

Note: we are using user-agent attributes from adobe to derive the device type (SMARTPHONE, DESKTOP, IOS etc)

what could be the reason of multiple user-agents associated with the same session id?
Thanks.

1 Accepted Solution

Avatar

Correct answer by
Community Advisor and Adobe Champion

Guessing that someone doing some testing on your website by changing the user agent from the browser's developer menu. Basically, all modern browsers have such a feature and user-agent is just a string from the browser which can be forged.

As long as you are not seeing a significant amount of such traffic, it should be fine.

View solution in original post

3 Replies

Avatar

Correct answer by
Community Advisor and Adobe Champion

Guessing that someone doing some testing on your website by changing the user agent from the browser's developer menu. Basically, all modern browsers have such a feature and user-agent is just a string from the browser which can be forged.

As long as you are not seeing a significant amount of such traffic, it should be fine.

Avatar

Level 1

Thanks @leocwlau 
Is there any alternative way to detect or derive the device type categorization from Adobe analytics feed data.
Any thoughts or visibility on this would be great..

Avatar

Community Advisor and Adobe Champion

Not really... 

 

User Agents are the way to identify a device, and emulating different device types is a standard way to test the site behaviour (breakpoint, specific logic based on device identification, etc) when you don't have a physical device.

 

I suppose you could try to assume that the first record is the "correct" one.. but emulation mode can persist on devices, so it's no guarantee... You could also maybe try comparing the device resolution to see if it scales with the screen size? If a user is on a desktop device but emulating a mobile device, the screen width and height should be a mobile aspect ratio but the recorded resolution might still be the full device resolution?

 

Or you could just make the assumption that if the same session has multiple identities that it's likely your testers and that you can treat it like desktop traffic... or exclude the data completely?