For a long time I assumed that most of my course sign-ups came from my Facebook posts. That was what the reports said, month after month. So I kept posting on Facebook, sometimes twice a day, putting real time into it.
A reader named Odette emailed me last year to say she had found my course through the Glenbrook Community Newsletter, a small local publication that had mentioned my work in their technology column. She assumed I had arranged it. I had not even known it happened.
What the referral data was hiding
The newsletter sends a plain text email with no tracking links. When readers clicked through to my site, there was no UTM tag, no referrer header that Analytics could read. Those visits landed in my reports as direct traffic, sitting quietly alongside people who had typed my URL from memory.
When I searched my Analytics account for sessions that landed on my course page and came from direct traffic, the volume was much larger than I had assumed. I had been ignoring a significant chunk of my audience because they were categorised incorrectly.
What I did after finding this out
I contacted the newsletter editor, Pieter, and asked if he would be willing to use a trackable link in future mentions. He agreed immediately. Within six weeks I could see clearly that the newsletter was sending more qualified visitors than my Facebook activity had ever managed.
The broader point is that direct traffic in Analytics is not one clean category. It is a catch-all for anything the tool cannot identify. Treating it as background noise rather than a signal worth investigating cost me over a year of useful information.