Bot Activity Tracking | Community
Skip to main content
C
Cameron_Craig
New Participant
April 20, 2023
Solved

Bot Activity Tracking

  • April 20, 2023
  • 1 reply
  • 2518 views

I enabled the Bot activity tracking just short of a month ago on our instance, but I am seeing numbers that seem to be lower than I would have expected, with a total of 330 bot activities identified: 

For reference, we have had around 240,000 delivered emails, 100,000 opens and 3700 clicks in that period. 

As a relative layman I'd take this outcome at face value and just chalk it up as a low impact on our numbers and get back to emailing, however we use a different tool for our transactional emails which seems to suggest a different story. 

 

Our Mandrill numbers show around 70% of opens are by the Google Image proxy email client, which I would have assumed is bot traffic since it is not a user initiated action. 

 

I have read some REALLY old discussions in the community that suggest that the Google image proxy triggers after the first open, which would be great news, but the Mandrill numbers suggest otherwise.

 

  1. To those with more experience, does our bot activity seem particularly low? 
  2. Is the Google Image proxy considered a 'bot' by the Marketo bot trackers? (can see a "Googlebot" on a sample of the IAB list, but I'd guess they have a few...)
  3. Is this even a Marketo issue, or more likely a Mandrill one?
  4. What other areas or issues could I look into to catch more bots?

Apologies for the long list of questions, I'd appreciate any insights you can share. 

This post is no longer active and is closed to new replies. Need help? Start a new post to ask your question.
Best answer by SanfordWhiteman

The Google Image Proxy (like any webmail image proxy) isn’t bot activity. It’s triggered by the client downloading images. It just so happens that the request is first made to the proxy, and then is forwarded to the origin server.

 

The Proximity Pattern numbers seem kind of low but then there’s no way to truly reliably detect mail scanners which are by their very definition, not supposed to be detectable. (If you can detect them, then a malicious server operator can detect them, too.)

1 reply

SanfordWhiteman
SanfordWhitemanAccepted solution
New Participant
April 20, 2023

The Google Image Proxy (like any webmail image proxy) isn’t bot activity. It’s triggered by the client downloading images. It just so happens that the request is first made to the proxy, and then is forwarded to the origin server.

 

The Proximity Pattern numbers seem kind of low but then there’s no way to truly reliably detect mail scanners which are by their very definition, not supposed to be detectable. (If you can detect them, then a malicious server operator can detect them, too.)

    C
    Cameron_CraigAuthor
    New Participant
    April 20, 2023

    I had thought that the proxy was used to pre-load the image in advance, and could have happened before a user initiated an action and potentially without the user ever actually opening the email. 

     

    Thank you for clarifying that and for taking the time to help. 

    Cameron Craig
    Terms & ConditionsAccessibility statement

    Loading footer...