Acoustic Campaign Ideas

The Acoustic Ideas portal has moved! Please sign up for our new Ideas portal and submit your suggestions at

Omit or Discount Improbable Activity Data

Enable users to suppress the email or website activity for a particular contact.

In reports and in raw data (Raw Recipient Data Report and Website Activity Report) we have observed contacts whose activity is improbable. Examples include:

  • "Clicking" every link in an email, including headers and footers, as well as hyperlinks only visible in the text version
  • Visiting the same web page 10-20 times in a 24 hour period
  • Visiting dozens of web pages multiple times, almost as if the contact were crawling the site

The cause of these reporting errors seems to be a bot, script, or spam prevention tool that systematically visits links on behalf of an email recipient.

These outliers tend to trigger false positive alerts, and skew reports. Scoring models that measure these activities become considerably less useful.

Omitting these addresses or their domains from mailings is not a viable solution since, in our experience, some of the clicks and visits are, in fact, valid.

There should be a way to either omit this outlier activity from reporting and scoring, or there should be more research to figure out the cause and report appropriately. Another solution may be a footnote or flag on a Contact Insight that says "The activity related to this contact may be inaccurate." 

  • Avatar32.5fb70cce7410889e661286fd7f1897de Guest
  • Aug 22 2018
  • Needs review
How will this idea be used?

This idea will be used to provide sales and marketing with correct data.

What is your industry? Other
What is the idea priority? Medium
DeveloperWorks ID
Link to original RFE
  • Attach files
  • Avatar40.8f183f721a2c86cd98fddbbe6dc46ec9
    Guest commented
    August 23, 2018 23:31

    We have noticed similar patterns and would actually like to isolate such subscribers and remove them from the database altogether. A report that helps us identify the activity described above, which is unlikely to be human, would go along way towards keeping our database clean of bots and spam.