The dashboard said 1,715 clicks. Thirteen of them came from a person.
That is not a tracking quirk or a rounding error. It is the normal state of cold email reporting. A security scanner opens every link the moment a message lands, so a link in an email sent to a corporate inbox gets clicked before anyone reads the subject line. The dashboard records it, buckets it with the rest, and calls the total engagement.
Report that number and your click rate looks healthy. Count the humans and it lands at 0.3%. The original post lays out the full breakdown, including the fact that three of those clicks were delivery checks and one was another link checker doing its rounds.
The inflation is not only in clicks
Opens get padded the same way. Gmail and other clients load images on their own, which fires the tracking pixel whether or not a person opened the message. In the campaign above, 2,097 opens were recorded and 123 of those were automatic image loads, leaving 1,974 verified. That is a modest gap in this case, but it moves in the wrong direction, and it moves more as volume grows.
None of this is malicious. Scanners and privacy proxies are doing their job. The problem is that most sending tools hand you a single blended number that mixes machine behavior with human behavior, and give you no way to separate the two after the fact.
Why rotated mailboxes turn a reporting bug into a strategy bug
A single mailbox with skewed numbers is annoying. A pool of rotated mailboxes and domains with skewed numbers is actively dangerous, because the dashboard is the only instrument you have for deciding which domains are healthy, and it is measuring the wrong thing.
Look at how the distortion lands across a portfolio. A domain that sends into corporate inboxes hits scanners on arrival, so it looks busy whether or not anyone read the mail. A domain that reaches people whose filters do not scan links shows only real clicks, so it looks quiet even when humans are engaging.
Now you have two domains. One looks productive and one looks dead, and the difference between them might be nothing more than the security posture of the recipients. So you keep a burned domain alive because the numbers look fine, or you retire a working domain because the numbers are quiet. Either way, the decision was a guess dressed up as data.
The ten-second question
There is one question worth asking any sending tool, and it takes about ten seconds to answer: of the clicks you show me, how many came from a person?
If that answer is not already in the dashboard, you are still guessing, and no amount of spreadsheet post-processing will rescue you, because the raw signal is gone. You cannot reliably subtract scanner traffic from a blended total. You can only measure it separately at the point of capture.
That is the entire argument. This is not a debate about attribution models or dashboard taste. It is a question of whether the number in front of you corresponds to a human being.
What a useful dashboard separates
The distinction has to live in the interface, not in a doc you wrote for yourself. The minimum is a separate column for machine activity, with clicks and opens split into recorded and verified. Scanner hits go in one bucket, automatic image loads go in another, and the number you make decisions on counts people only.
A few things worth checking when you evaluate a tool:
- Does it label scanner clicks separately, or fold them into the total?
- Does it distinguish automatic image loads from real opens?
- Can you see the verified count at the domain level, not just the campaign level?
- Does your domain health view use verified engagement or the raw number?
If the answer to the last one is the raw number, your rotation logic is running on noise.
Mailheight builds the split into the dashboard. Scanner clicks and automatic image loads get their own column, and the verified figure counts people only. On the campaign above, that was 13 out of 1,715. That is a very different story from the one the raw total tells, and a far more useful one.
Thirteen clicks is not a disaster. It is information. You can work with thirteen real clicks. You cannot work with 1,715 clicks that mostly came from a machine, because every decision you make from that number, which domain to keep, which subject line to scale, which segment to push harder, is built on a fiction.