What a link is worth on a page with no traffic

A buyer sent me $250 for a link on a page that will be read by almost nobody. I knew that when the money landed. So did he, in the sense that the information was available to him, on the open web, in about thirty seconds.

He never looked. He had three numbers about the domain and none at all about the page, which is how nearly every one of these deals gets done.

I have been on the other side of it too, so this is not me being clever at a buyer’s expense. The question underneath the transaction is real and most people answer it too fast in both directions.

You bought one of the two things a link does

A link passes something to a search engine, and it moves a person from one page to another. Two jobs, one price.

Every rate card, every media kit and every negotiation I have ever had prices the first job. The second one is treated as a rounding error, or more often it just never comes up.

On a page with no audience the second job does not happen at a reduced rate. It happens zero times, permanently, on a page that will be crawled by machines and read by nothing with a pulse.

Why “worthless” is the lazy answer

The satisfying line here is that a page nobody reads is worth nothing. I do not believe that and I would rather be accurate than punchy.

Crawlers do not need readers. The link gets fetched and counted in whatever the graph counts, in the same way as a link on a page with ten thousand monthly readers. Pages with no human audience have moved rankings for years and still do, and anyone flatly telling you otherwise has not tested it.

So there is real value in the crawl half. What you have done is buy half a product at the price of a whole one. That is a narrower complaint and a truer one.

The half you lost is the half that survives

This is the part that decides it for me.

Ranking value is rented from a company that changes its mind on its own schedule, several times a year, without warning anybody. Every operator has watched a page that worked stop working, with no explanation on offer.

A click is different. Somebody arrived, read something, and either bought or did not. That value landed in a bank account and stayed there, and it verified itself on the way in. No update reverses it.

A zero traffic placement gives you all of the volatile half and none of the permanent half, priced as though you received both.

Four things that decide whether a quiet page is worth anything

Quiet pages are not one category. These are the checks, roughly in the order they change my mind.

Whether a person can reach the page at all. If the only route to it is a sitemap entry or a crawler that already knows the URL, the site itself does not treat that page as part of the site.

Whether anything else on that domain gets read. A quiet page on a live site is an archive. A quiet page on a site where every page is quiet is a warehouse, and you are renting shelf space.

Whether the page sits in a section the publisher treats as real. Same template, same navigation, same author byline as everything else, or dumped in a directory called resources that nobody has touched since it was created.

Whether anything outside the site points at the page. If nobody has ever linked to it, there is very little arriving for your link to take a share of.

The click test

The first one takes under a minute and it disqualifies more placements than the other three put together.

Take the URL the seller gave you. Open the site’s homepage. Now try to arrive at that page by clicking, the way a reader would. Down the navigation, into the blog, through the categories, into the archive. Three clicks.

If you cannot get there, the page is an orphan. It is a file with your link in it and no route by which any human reaches it.

There is a middle case worth naming, because sellers live in it. The page is technically reachable, but only via a category listing five levels down that the navigation never surfaces either. The site links to it the way a warehouse links to a box at the back. Treat that as an orphan with paperwork.

It is a browser and half a minute, and most people buying links have never once run it.

The question that reprices everything

Would you pay for this link if the metrics did not exist?

No authority score. No traffic estimate. No referring domain count. Just the page, and the sentence your link will sit in.

If the answer is no, the link is incidental to the purchase. What you are buying is a metric.

And metrics are the most manufacturable object in this market. A score computed from a link graph rises when a site acquires links, which is precisely the trade a link seller is in, so the number on a seller’s media kit is partly a product of the selling. Traffic estimates are models. I can hold those models against real analytics on properties I own and they come out wrong by a factor of several in both directions.

So the thing you are paying for is the easiest thing in the whole transaction to fake, and the thing you could confirm with your own eyes in half a minute is the thing you decided not to check.

Price it as if the tools shut down tomorrow

The replacement question, and the one I actually use: what would this placement do for me if every third party SEO tool went dark and never came back?

Would a person in my market ever land on that page. Would I be glad to have my name on it if a potential customer read it. Would it produce one enquiry in a year.

If those are all no, the remainder is the crawl signal on its own, and that number is a great deal smaller than the quote. I have written separately on how to price a placement from the seller’s side; this is the narrow case underneath it.

My number

I will not pay more than a quarter of asking price for a placement on a page I cannot reach by clicking from the site’s own navigation.

In practice that means I do not buy them, because no seller discounts by 75% and we both know why.

The placements I bought last year came to $387 across three sites. Exactly one of the three sits on a page a real person could plausibly arrive at. I knew that about all three before I paid, so the rule I just gave you is one I broke twice with my eyes open. That seems like the right amount of respect to have for it.

The quarter figure has no research behind it. It is roughly what I think the crawl only half of a link is worth once you strip out the half that brings a customer. Land somewhere between a fifth and a third and I would not argue.

The part where I am the problem

I sell dofollow placements on my own sites at $250. Some of what I sell is exactly what this article describes.

There are pages on my properties that essentially nobody reads, and a link on one of them is worth less than the invoice says. If a buyer asks me directly, I tell them. Across everything I have ever quoted for, the buyers who wanted to know which page before paying can be counted on one hand.

When I audited outbound across my properties this year, the count came back at 15,321 dofollow links leaving against 16 coming in. That describes my own inventory, not somebody else’s. It tells you roughly what any single link leaving that property carries, and I have sent invoices for links leaving that property.

What changed is small. When somebody asks for a placement now, I point them at a page the site genuinely links to, for the same money, instead of the quiet one that would have been easier to hand over.

What the seller tells you by which page they name

Ask for the exact URL before money moves. It is an ordinary request and the reaction is worth more than the answer.

A seller who names the page immediately has a page. A seller who says they will place it on the most relevant article is telling you they are going to write one. A seller who will not show you the page until after payment has already answered it.

When a quiet page is genuinely fine

There is a legitimate version of this and it deserves saying.

A real publication with a deep archive has thousands of pages nobody read this year. That is what an archive is. A five year old article on a site with a working newsroom is quiet for the same reason an old newspaper is quiet, and it belongs to a body of work that gets linked to and crawled constantly. A link inside it inherits some of that.

So the test is never whether the page gets traffic. It is whether the site has anything at all that does.

What I do not know

I have no discount curve. I cannot tell you a zero traffic placement is worth 30% of a trafficked one, because I have no dataset that says so and neither does anyone quoting you a number.

I have also bought exactly this kind of link knowingly, more than once. The reason was that I wanted a link from that specific site for what the rest of the site was, and the page they offered was the page they had. I judged the site worth paying for while knowing the page carried nothing.

Buying a dead page for that reason is defensible. Buying one because a tool scored the domain well is not, and that is why almost everybody actually pays.

Negotiate the page, not the price

Run the click test before you open a single dashboard. Then ask whether you would still want the placement with the score removed from the pitch.

If the honest answer is no, stop haggling over the number and start haggling over the page. Ask for a placement inside something the site already links to from its own navigation and already sends people to. Sellers agree to that far more often than they discount, because it costs them effort instead of revenue.

If they refuse, you have learned what the site is for free, before spending anything. Everything else I have worked out from both sides of these deals is here.