best way to identify bot clicks?

Highlighted
Anonymous
Not applicable

best way to identify bot clicks?

Hey all,

I've read a few discussions here about bot clicks inflating email metrics, which I think I'm observing as well. What's the best way to filter out these bot clicks in reporting?

I've seen some suggestion on creating a smart list and use both an email open and a email clicked filter. But doesn't this also over count if the behavior is click by bot, open by person but he/she never clicks?

the other option I see is adding in a hidden link that a normal person wouldn't click on. and use smart list to filter these out. For this method, do you just add a <a> tag somewhere high in the email so that it's a first link and make it blend into the email background?

3 REPLIES 3
Highlighted
Level 10 - Community Moderator

Re: best way to identify bot clicks?

The granularity of Marketo Smart Lists doesn't let you to filter out mail scanner clicks reliably, as noted in other posts.

You can get somewhere with SLs, but advertising more than partial coverage to higher-ups would be misleading. And, as technology is upgraded, it (rightfully!) gets harder and harder.

Ultimately, such detection needs to be done in Marketo's core tracking infrastructure to have any change at working. It'll still have gaps but it's the only way. Or you can add logic to your own tracking server... we run our own tracking server for a few clients but haven't gone this far as it really is a complex issue,

Highlighted
Anonymous
Not applicable

Re: best way to identify bot clicks?

Thanks Sanford, your right that at best it's partial truth but that feels better than looking at raw click numbers and be completely mislead on email effectiveness.

What in your opinion is the best immediate solution through marketo that gives reasonable confidence in that data?

Highlighted
Level 10 - Community Moderator

Re: best way to identify bot clicks?

Unfortunately, none of the attempts have been suitable for my professional recommendation, and none provide reasonable confidence.

I think it makes more sense to concentrate on Visits Web Page events and start treating Clicks Email as merely directional (that is, if you see an uptick or downtick in clicks for the same email, if you can't attribute that trend to updates to/introductions of mail scanning technology, then treat it as a trend in engagement).