The Pages Nobody Reads Are the Ones Trying to Do Everything

I keep the page views for my own site in a database I can query. Not a dashboard somebody rents me. The actual rows.

This week I stopped asking it how many people came.

I asked it how far down the page they got before they left.

The answer is the most useful thing my own site has told me all year, and it is the same thing I would tell any business owner who hired me to look at theirs.

How the numbers were counted

Seven days ending September 12, 2026. Before counting a single visit I threw out every hit the site tagged as automated, two known synthetic browser strings that fake being a phone and a Linux desktop, and every session that never scrolled a pixel.

What is left is people. It is a smaller number than the one the dashboard would have shown me. It is the only one worth acting on.

How far people actually read

Page	Human views	Median scroll
/case	51	17%
/ (the homepage)	41	2%
/the-story	20	64.5%
/book	17	22%
/fuel	15	83%
/book/preorder	15	24%
/the-map-room	12	29%
/j6	5	100%

Where they land, and whether they stay

Landing page	Sessions	Avg pages per visit	Left after one page
/	27	2.44	20 of 27
/case	17	2.35	11 of 17
/book	12	2.50	8 of 12
/book/preorder	12	1.17	10 of 12
/the-story	6	4.17	0 of 6

Read that again

My front door takes more landing traffic than any other page on the site.

The middle visitor reads two percent of it.

Twenty out of twenty seven leave without ever opening a second page.

Now look at /the-story. Six people landed there in seven days. The middle one read sixty four percent of the page. Not one of them stopped at one page. Zero out of six. They averaged four pages a visit.

Same site. Same week. Same traffic sources.

One page loses seventy four percent of the people who touch it. The other one loses nobody.

The thing the two lists have in common

I sat with it for a while and the pattern is not subtle.

Every page that holds people does exactly one job.

/the-story tells one story, start to finish. /fuel makes one ask. /j6 is one index that does one thing, and the median reader goes all the way to the bottom of it.

Every page that loses people is a menu.

My homepage is a menu. The case hub is a menu. The preorder page asks you to pick between three things before it has told you why you should want any of them.

A menu is a page that makes the visitor do the work of deciding what this is. Two percent of a page is about how long it takes somebody to figure out they are being handed that job, and hand it back.

What this does not prove

Six sessions is six sessions. /j6 at one hundred percent is five views. These are small samples pulled at a low point in my traffic, and I am not going to pretend otherwise.

The single-job versus menu read is my conclusion off my own numbers. It is not a law. Somebody could land on /the-story already sold and read all of it for a reason that has nothing to do with page design.

I also cannot tell you these visitors are the same kind of person. Somebody who clicks through to a story from a post is warmer than somebody who types my name into Google.

What it does prove

On this site, in this window, the page taking the most traffic is the page nobody reads, and the pages with the deepest reading are the ones that do one thing.

It is not a traffic problem. Forty one humans reached my homepage. They arrived. They just did not find a reason to go down.

That distinction matters, because "get more traffic" is expensive and "give the page one job" is free.

What I am doing about it

I am not redesigning the homepage this week. I am rewriting what it is for.

One job above the fold. One line that says what this place is. One next step. Everything else moves down or moves out.

Then I measure the same number again in seven days, and if the median does not move, I was wrong and I will publish that too.

If you own a business, this is your homepage

You have the same page. Almost everybody does.

Yours has a slider, a services grid, a testimonial row, a map, three phone numbers and a chat bubble. It looks finished. Your web guy is proud of it.

And the middle visitor is reading the top two percent of it and leaving.

Here is the test, and you can run it today without buying anything. Open your homepage on your phone. Do not scroll. Read only what you can see.

Does it say what you do, who it is for, and what to do next?

If it does not, the rest of the page is decoration for people who already left.

The fix is not more pages. It is fewer decisions.

One job per page. One next step per page. Say the thing before you show the menu.

That is most of what I do for people, and it is why I would rather look at your numbers than your design. The numbers tell you where the leak is. The design is just where you go to plug it.

If you want somebody to read your own rows out loud to you and tell you what they say, that is here: https://www.realryannichols.com/services

The record

This is my own database, my own site, my own bad number published on purpose.

I do not want anybody taking my word for how this works. I want you to go pull your own scroll depth and look at it.

Do not threaten anyone. Do not harass anyone. Do not contact anyone in my name.

Read it. Share it. Send receipts, evidence, witnesses. Put eyes on the record.