Spam filters registering clicks?

[SanfordWhiteman]

filter service-only string to their UA when checking links

Doesn't make sense though.  That would make their service worthless because it's supposed to be prechecking for hostile sites.  All a site would need to do is UA sniff and return a non-malicious payload.  They need more randomness, not less (my experience is it isn't the same UA at all).

[Beth_Rotach]

I'd be happy to reach out to Barracuda - that's the culprit for us about 90% of the time. What do I need to do/say?

[SanfordWhiteman]

"Please don't protect your clients against phishing and malware attacks"?

Seriously, I don't see why any anti-spam software that offers this functionality in the first place would offer to turn it off.  It's obviously by design that it follows links, even if it seems like a fiasco from an analytics standpoint. It might also be better to not contact them because according to the research here, there is a way of avoiding spurious clicks, but they could break that if they wanted to (by starting to follow JS redirects).

[Anonymous]

One of my big concerns is that we're passing this activity into SFDC, which our inside reps are using to follow up on what seemingly look like responses to email campaigns. Have others resulted to just disabling that type of activity from being passed to SFDC?

[SanfordWhiteman]

One of my big concerns is that we're passing this activity into SFDC, which our inside reps are using to follow up on what seemingly look like responses to email campaigns. Have others resulted to just disabling that type of activity from being passed to SFDC?

If you mean replicating all activities (well, one per lead per activity type per day, technically), yes, you would have to turn that off if you want to perform any filtering.  If you create a Smart Campaign that seems to work, like the one Conor proposes above, you can use that to create SFDC tasks and/or Interesting Moments that correspond to the filtered activities/sequences of activities.

[Anonymous]

Marketo support suggested we trigger off of Opens Email and Clicks Link in Email, but the issue we're seeing is that this is still generating false positives, because those if a company has the filter, they're still going to register a click. So if they even just open an email, it looks like they've satisfied both requirements. In many cases, the clicks are even logging far ahead of the time when someone opens.

Has anyone worked out a way to tell Marketo that the Email has to open BEFORE the click?

[SanfordWhiteman]

It's kind of reckless to assume a deep scanner won't download images (not to say they all do, but there's no reason for them not to).

But if you want to consider the Open to be human, then filter your Click triggers by a previous Open.

Still, I'd say this is an even more fragile variant of the Visit Web Page filter above, which isn't fully satisfactory.

[Beth_Rotach]

Just a quick note - we've been emailing pretty aggressively with Marketo support regarding this issue. We also found out that MANY other ESPs provide this "click filtering" as part of their service because it happens so often to folks. We recently spoke with about 8-10 other ESPs that automatically (and very easily) filter these clicks for their clients. Apparently the ESP can easily filter clicks by IP/known barracuda IPs and code. Marketo deliverability team assured that their are now working with the product side to try to implement this ASAP, especially because so many people are asking about it - KEEP ASKING!

[SanfordWhiteman]

Apparently the ESP can easily filter clicks by IP/known barracuda IPs and code.

Nope, they cannot (and are not) doing it this way. It's a preposterous claim, and any ESP that claims to reliably filter automated clicks this way is lying.

Rather, they are using a mechanism that is closer to what I have described in this thread and elsewhere.  It is possible for Marketo to attempt the same and achieve a high degree of coverage.  But to the degree that it works, it is because of the defensive coding used by the mail scanner (to prevent an amplification attack against the scanner itself) and not because of any special brilliance or detective work by the ESP.

[Robb_Barrett]

OK, here's a situation I was presented with this morning:

We use Marketo for our Contact Us page and we have workflows that fire off alerts. One of the alerts has two links: I've Handled This, or I Need to Re-Route This. There is a follow up workflow that is triggered on the I've Handled This click.  One of my colleagues asked for help because a click is firing off the follow-up workflow 4 times.

One of the logs I looked at shows the initial alert delivered Sunday night.  On Monday morning, the alert was clicked at 10:34am and there was a corresponding VWP.  Then, also at 10:34 I see two more Clicks Link and only one VWP.

My first thought is that I have a double-clicker. I created a lead for myself following the process.  I was very slow about waiting to click on the link the first time. I did, it registered one click, one VWP, then nothing more. I put in a filter for Not Clicks Link In Email in Past 1 minute to see if that would help.  Then, about 5 minutes after my first click I double clicked. It registered two clicks and two VWPs.

A minute later, it registered 3 clicks and 1 VWP. These were not by me or anyone else.

Now, it's work noting that we have a URLDefensePoint system in place. All links in emails are re-coded by the server with DefensePoint to check. I'm thinking that it's testing the link for us to see what happens prior to allowing the browser to go to the link.

Thoughts?

Sanford Whiteman​