Nuances of Email Tracking: Understanding Genuinity of Open and Click Activities in Marketo

Marketo uses tracking pixels to track email opens. However, nuances arise when distinguishing genuine human engagement from automated bot activity. Let’s delve deep into those nuances today!

The Challenge of Tracking Opens

Marketo (and other MAP/ESP platforms) rely on tracking pixels embedded in emails to register open activities in the recipient's Activity log. Yet, this method encounters a hurdle—many email clients block image downloads by default. Consequently, if a recipient interacts (e.g., clicks a tracked CTA) with an email without downloading images, Marketo may not detect an open activity (see the snapshot below from Outlook’s desktop application).

Many users would consider such activities (i.e., clicks w/o opens) as those coming from the bots and thus not genuine; however, this is not always the case. This poses a challenge as genuine human engagement may go unnoticed.

The Click-Open Discrepancy

Moreover, the sequence of actions complicates the interpretation of email activities. For instance, a recipient may click on a call-to-action (CTA) before opening the email, bypassing the traditional open activity. While the Email Performance report backfills open activities for recipients without one but with a click activity, this data isn't reflected in the recipient's activity log.

Analyzing Activity Data

Additionally, Marketo's smart lists lack native functionality for time-based activity comparison in case users want to compare the time difference b/w the sent, delivered, open, and subsequent click activities to analyze the genuinity of activities. As a workaround, exporting bulk activity data enables deeper analysis.

However, identifying instances where clicks precede opens doesn't conclusively indicate bot activity. Recipients may download images after clicking the email, skewing the sequence of events (you’d see clicks before opens in that case). However, Marketo’s proximity pattern recognition bot detection method also classifies the bot activities based on the proximity of activities logged.

Lastly, I’d recommend leveraging behavioral cues and contextual data to gauge user engagement (e.g., form fills, interactions/engagements on social/other channels) rather than just relying on email performance data.

Happy Marketo’ing! 💜