[Shared Blog]: I Want to Believe, But: Your Email Link Clicks Aren’t Real

[Dan_Stevens_]

I recently came across this external blog post by Courtney Grimes​ and thought it would be useful to share with the community (great post, btw, Courtney).  Especially since there has been a lot of discussion/questions recently on false-positives of email opens and clicks:

https://www.demandlab.com/insights/blog/want-believe-email-link-clicks-arent-real/

I did want to comment on the guidance to combat this:

After testing this idea on dozens of different Marketo instances, the single most effective way I’ve found to track email clicks is a combination of two filters: Email was delivered + Visited web page: [web page linked to in email]

Typically, one would simply add these two filters to the smart list of the trigger campaign.  When we tried that in the past, it didn't work for us since Marketo doesn't log this activity immediately as it happens - at least not in our case (especially the "visited web page" activity).  We've seen up to a 45 minute delay.  So we used an alternate approach: we included a "visited web page" choice in our "change program status" flow step.  And prior to that flow step, we also included a 45 minute wait step.  The only issue with this approach is if the user ever visited this page in the past, the choice will still be true.  In other words, the choice step is not directly tied to this instance of the email click.

As Courtney states in her wrap-up, there really is no fool-proof solution.  But if you had to pick one, this is probably the best approach.

[Marketo_HyperX_]

Our Customer Support Manager Mariska noticed that the clicks happen 1 minute before the email is delivered.  Did you all see the same behavior?

To avoid wait steps, do you think this would work:

Thanks!

Amy

[SanfordWhiteman]

Your Smart Campaign doesn't reflect the order your describe. Please look at it more closely. Also: give up on trying to solve this using Smart Campaigns.

[Marketo_HyperX_]

Hi Sanford.  The false click happens 1 minute before the email is actually delivered.  Thus, I believe my smart campaign logic is correct and will filter out the false clicks.  Thoughts?  

[SanfordWhiteman]

Smart Campaigns do not allow for 1m granularity.

Again if it were this easy, no one would be worried about it. In fact all you're seeing is some of the mail scanner clicks. The other ones are undetectable. That's the whole idea.

[Michael_Guanci]

So far we have had decent results using a combination of Dan's method and Sanford's redirect.  However, I noticed that some members from eBay are getting counted as a real click even though they all happen within a minute after the email is sent/delivered.  They are clicking the link and visiting the page that has the re-direct.

This is for our ABM program so we want to include eBay in this moving forward (we don't want to exclude them is what I am saying), is there a time constraint I can add that I am not thinking of?  Like did not click on email within 3 minutes of being sent?  I am following the guidelines Dan outlined for the 30 minute wait and seeing if they visit the web page. 

[Grégoire_Miche2]

Vote here:

[Casey_Grimes]

A quick update here: as of today, I've picked up some new behavior by a different major antispam provider that may cause further wrinkles in this, especially around user agents (because I'm detecting different things in Marketo and on the server itself. :/) This particular company is using Amazon EC2 instances to check links and is making it to full web page visits. Hoping to have more clarity on this in the next week or so.

[Kevin_McMahon1]

So I've read through each comment and am still confused on the current best method to combat false positives.

Is it Dan Stevens​' approach using the Visited Web Page flow? Is it using utm parameters on every link?

Sanford Whiteman​ I'm also curious about the method you're testing to track the user clicks vs automated clicks.

[SanfordWhiteman]

Sanford Whiteman I'm also curious about the method you're testing to track the user clicks vs automated clicks.

I'll surely get to documenting it but I have an absurd amount of projects running at the same time, and blogging is hard, gotta get your diagrams tight and everything.

On the other hand, walking you through it live is pretty easy, so if you want to DM me I can show you what it is.

[Casey_Grimes]

I wanted to go ahead and piggyback on this particular thread, as there's a new complexity at play: an anti-spam measure has been updated by at least one company in the past couple of weeks to fully load pages in emails, including with JavaScript. Right now, because the IPs looking at this are Azure-based, it's a straightforward process to remove them by not loading Munchkin on Azure IPs and/or screening those IPs out of Marketo activities, but this sort of development is worth keeping an eye on.

More information is at https://www.demandlab.com/insights/blog/want-believe-butyou-cant-coast-anti-spam-measures-year-ago/

[SanfordWhiteman]

This has been the case for awhile, as if they weren't running JS they wouldn't be able to follow tracked links to start with. They're headless browsers after all.

[Casey_Grimes]

Yes--the difference now is that they're using headless browser sessions instead of just doing DNS resolution, essentially. It's been pretty uncommon prior to this new wave of activity.

[SanfordWhiteman]

It was never just DNS, though. They followed the tracking redirect using JS. (And sometimes ran JS on the next hop, too.)

[Dan_Stevens_]

Sandy, check out Courtney's new blog post.  Does this explain the multiple clicks/visits activities that were being captured in our test lead record the other day?  For what it's worth, we're in a O365 environment at Avanade. 

It might be also helpful to point out that every link in our Outlook environment is wrapped in a "safelinks.protection.outlook.com" URL (with the actual URL referenced as a URL parameter).

[Anonymous]

wrote this a while ago: Marketers vs email security bots: an emerging war | B2B Lead Nurture

would be nice if Marketo formally addressed this in its backend, or by opening a backchannel to the leading bot security vendors to negotiate whitelisting marketo servers.

[SanfordWhiteman]

....backchannel to the leading bot security vendors to negotiate whitelisting marketo servers.

The mail security vendors should never even consider this. Since Marketo doesn't do any egress filtering, you can send any link, including a link to a phishing site.

[Abby_Ryan]

Hey All,

I just wanted to chime in with another solution that's been working for us for a while. We, too, noticed a large uptick in the number of false clicks about two years ago. One of the reasons behind this was that when I first started here I worked with IT to have the SPF and DKIM set up properly, which had a positive impact on our delivery (~97% up from ~70%), but because we were able to 'hit' more companies now, we were also encountering more spam filtering software. As machine learning continues to grow and expand, we knew we needed to draw a hard line under how we track our emails.

We now use the "Visited Web Page" trigger, with Querystring Parameter Constraints built in every time. We utilize special UTM parameters (UTM_source, UTM_medium, UTM_campaign, etc.) to delineate visitors to the page who came from the email. It's important to set these Querystring constraints to use "contains" and NOT "equals". This is because the string will not only contain the UTM, but also the user's munchkin code, etc. and you don't want to lose data. See screenshot below for an example.

To make sure that we're not double-processing or double-scoring any lead, we also exclude the email's UTM from the overall web campaign that's tracking the asset on it's own (see screenshot below).

Overall this is a bit easier to track once it's set up, and a huge benefit is that it keeps you from having to rely on long wait steps, which can also push your campaigns further down the campaign queue in priority and can lead to race condition errors.

I hope this was useful!

Thanks,

Abby

[SanfordWhiteman]

... but that doesn't refer directly to links in emails (I mentioned above that if you can UTM-tag all your links, this can be a near-substitute for click tracking, but it can't be complete).

[Dan_Stevens_]

Especially when you start leveraging some of the enhanced techniques to optimize the user experiences that have been shared here - like Sanford's "direct-download-and-apply-cookie-without-landing-page" technique: http://blog.teknkl.com/stop-using-direct-download-links-unless-you-like-losing-tracking/

I have to admit though, I really wish there was some stronger support with native UTM management in Marketo so that we can apply UTM parameters at scale within all of our activities - especially email.  Again, there's been some good discussions/hacks on ways to do this today (e.g., Marketo Hack: Automatically add utm parameters to Marketo emails - Erik Heldebro ), but not a scalable/integrated approach.

[Abby_Ryan]

Gotcha - sorry Sanford. This works for our purposes, but obviously not for all.