Twenty visitors in June. Two hundred and thirty four in July. We didn't run an ad, submit to a directory, or email a list. We built a homepage, wrote a handful of blog posts, and posted the headers on X and LinkedIn.
Then we opened Clerion on Clerion and asked it what happened.
The channel that grew the most has no logo next to it in any referrer table. It's labeled "Direct."
Your best channel isn't the one with a logo next to it in your analytics.
Our traffic went from 20 visitors to 234 in a month, and the biggest source doesn't show up as a referral
Page views: 1,640, up 712% on the month before. Visitors: 234, roughly eleven times the 20 we had the previous period.
Direct brought in 351 of those visits, 67.4% of everything, up by 292 visitors month over month. That's bigger than X (71 visitors, 13.6%), Google (63, up 53) and LinkedIn (30) combined. Direct isn't nothing and it isn't unlabeled traffic hiding a mystery. It's word of mouth without a trail. Someone reads a post on X, doesn't click, and types the URL in that evening, or the next week, from a note they left themselves.
Every blog post gets written as five headlines, and paragraphs get added reluctantly
Nothing goes to draft until the section headers exist. Each one has to survive alone: if we wouldn't publish a header as its own post on X, the section under it is weak, and we fix the header before we touch the paragraph. Four to six per post, never more. One line per post gets to carry the whole argument. One metaphor, used once.
Two posts written this way, "Your numbers now count people, not bots" and "Analytics for people without a data team," are still pulling entries weeks after publishing, 16 and 30 visitors this period alone. Neither was promoted again after the week it shipped.
The blog didn't spread. The headers did, republished exactly as written
Every post ships with the same four things every time: each header lifted whole onto X, a compressed line for LinkedIn, one contrarian angle for replying under someone else's launch post, one useful line to leave under a founder's complaint about their dashboard. Nothing gets rewritten for the platform. If a header can't survive alone on X, we rewrite the header, not the tweet.
That's why t.co shows 71 visitors and LinkedIn shows 30, small numbers next to Direct, but they explain Direct. Someone screenshots the header, forwards it in a DM, three days pass, and a visit lands with no referrer attached to any of it. The post that gets credit in the dashboard isn't the one that did the work.
The visitors came back fast. They didn't stay: 27 minutes of attention collapsed to under seven
Average session dropped from 27 minutes to 6 minutes 49 seconds. Bounce sits at 69%. Mobile is a fifth of sessions and leaves faster than desktop does.
That's the actual trade of this month, not a footnote. More people are arriving, looking at one page, and leaving. Fast, shallow traffic and slow, deep traffic are not the same win, and reporting only the first number is how a real change gets mistaken for a good month.
Twelve months of this unchanged looks the same as this month, just louder: another spike from another good post, another 69% bounce rate, another exit before the second page. The number on the dashboard keeps climbing. The number that pays the bills doesn't move with it.
What to check this week, no tool required
- Open your analytics and find your top referrer. If it's Direct, don't shrug at it. Search your last three posts on X or LinkedIn by URL and see which ones got shared or screenshotted. That's your unlabeled channel, and it has a name even if your dashboard doesn't give it one.
- Put your bounce rate next to your top entry page. Above 60% with no next step on the page (no link, no related post) means the page is a dead end. One line at the bottom pointing somewhere fixes the easiest half of that number.
- Compare your visitor trend to your session-length trend over the last 30 days. Growing visitors with shrinking time on site is a distribution win and a content problem, stacked on top of each other. They need different fixes, and treating them as one metric hides which one is actually broken.
Direct traffic is an echo, not a click. You post somewhere else entirely, and days later a stranger walks in the front door with no memory of where they first heard you.
We don't have an ad budget, so word of mouth doing the heavy lifting is the whole plan. If you want the same view of your own numbers, it's free to start and no card is asked for.