How do you identify fake engagement or pixel inflation?

Real engagement follows human rhythms. Fake engagement follows machine rhythms. Once you know what to look for, the difference is usually obvious.

Signs of bot-driven click inflation:

• Clicks arriving within 1-3 seconds of message delivery (no human reads and clicks that fast)
• All links in the email clicked simultaneously, in the same second
• Traffic from IP ranges associated with data centers and cloud providers rather than residential ISPs
• Generic user agent strings like "python-requests/2.28" or crawler-style agents
• Consistent click-to-open ratio of 100% or higher (every "opener" also "clicks")

Signs of MPP-inflated open rates:

• Open rates above 70-80% for lists that aren't highly engaged newsletters
• High open rates concentrated in Apple Mail device type
• Opens clustering at the exact time of delivery rather than spread over hours
• Open rate and click rate diverging sharply (40% opens, 0.5% clicks)

Your ESP may give you a head start here. Klaviyo flags machine opens separately. Mailchimp has introduced MPP filtering. If your ESP segments machine vs human engagement, use that filtered view for any real decision-making.

For the deep dive on why spam filters click links automatically, that explains the security-scanner side specifically. And if you want to understand how Apple's MPP works mechanically, that covers the proxy caching behavior in more detail.

The short version: if a click came from a data center IP within two seconds of delivery, don't let it trigger an automation. Most modern ESPs let you filter these out if you know where to look in your event settings.