I compare soft-launch traffic week-over-week with UTMs, not vanity sessions

I used to open Analytics after a soft-launch week and feel busy. Sessions up. Bounce weird. Referral blob said "Direct" for half of it. Zero idea which board or directory actually sent someone who stayed.

So I started tagging every outbound link with a tiny UTM before I hit submit. Not a fancy naming system. Just `utm_source=boardname`, `utm_medium=softlaunch`, `utm_campaign=weekN`. Same week, same campaign tag, different sources. Then I look at the week after and ask one boring question: which source still shows up in signups and replies, not just sessions.

Week 3 of one soft launch taught me this the hard way. A daily board dumped ~180 sessions in two days. Looked huge next to a quiet niche directory that sent maybe 40. The board people bounced in under twenty seconds. The directory people? Three of them emailed within 48 hours asking a real product question. Sessions lied. Replies did not.

I keep a one-tab sheet now. Columns: source, sessions, unique signups, stranger emails, paying or trial starts. I ignore "engaged sessions" when the product is early. I care if someone from that UTM asked something I could answer without a sales script.

Vanity is easy when you run Aura++ and watch founders celebrate a spike that never turns into a conversation. Soft launch is for learning which channel matches the buyer, not collecting a pretty chart. When I am picking where to spend the next week I still skim Best Launch Platforms, but the decision comes from last week's tagged rows, not from the board with the loudest homepage.

I wrote more about staying discoverable past day one in what happens after you launch. Same idea — compound the channels that leave a trail, drop the ones that only inflate sessions.

If you soft-launched in the last month, which UTM source still produced a human reply after the spike died?