Auditing the link profile of a site you just bought

The first thing some new site owners hear from me is an email explaining that they have inherited a deal with a stranger.

I sell placements. When a site carrying one changes hands I find out from a crawl rather than from a person, and then I write to somebody who has never heard of me to say there is a paid link with a term attached on an article they have not read yet. My only evidence is a ledger I keep myself.

That email exists because the link profile is the one part of a website acquisition that nobody hands over properly. The content comes with a database. The traffic comes with an analytics login. The links, which are usually the actual reason for the price, come with a screenshot of an authority score and a chart the seller has never interrogated.

What you are actually buying

You can rewrite the articles. You can rebuild the traffic across a year of publishing. You cannot go out and reacquire that link profile for anything close to what you just paid for the whole site.

So the diligence should be weighted the way the value is. In practice it goes to whatever is easiest to look at.

Live now, not live in an index

Every backlink tool reports what its crawler saw last time it went past. Links stay in those indexes long after they stopped existing on the page, and on small sites that lag runs to months.

So take the referring domain export and verify it yourself. Fetch each page, search the raw HTML for a link pointing at the domain you are buying, and record present or absent. The source, not the rendered page you skim in a tab.

The count that comes back is lower than the listing. I have never once run this exercise and found more live links than the tool claimed.

If the list is too long to check exhaustively, sort by whatever authority measure justified the price and do the top fifty by hand. That end of the list is where the money went.

While you are in there, note where each live link lands. Sites lose content, and a cull or a migration leaves part of the inbound profile pointing at URLs that now return nothing. A redirect map fixes most of that in an afternoon, and it is a fair thing to raise in the negotiation, because the seller has been counting links that go nowhere.

This is a snapshot and it should be. What happens to those links over the following years is a separate question. Today you need the number that exists on the day you pay.

Which ones was somebody paying for

Some share of any respectable profile was bought. A portion of that is recurring: a monthly retainer, a directory listing that renews on a card the current owner is holding, an arrangement that survives only while the relationship does.

Those links leave with the seller. Either the payments stop, or the host hears the site changed hands and resells the slot.

Four questions, one email. Which of these did you pay for. What did each cost. Is anything recurring. Whose card is it on.

Most sellers will not have an answer, which is information in itself. Somebody who cannot say which of their links they bought has not run the site the way the listing implies. If they do have records, ask for twelve months of invoices rather than a summary. A charge on a statement is harder to be vague about than a memory.

How much of it comes from the seller’s own hand

This one gets past people who did everything else properly.

Operators own portfolios. The person selling you one site often owns several more, and those have been linking here for years because it cost nothing.

After the sale those links come down, or they stay and count for much less than the export implies, because they were always one hand writing to itself.

Look for the pattern before you pay. Shared hosting fingerprints, an identical theme, the same author bio, sites appearing in each other’s articles, matching registration details. If a real share of the inbound profile comes from properties the seller is keeping, you are buying a number that exists only while they feel like leaving it there. Get it in writing or take it off the price.

The shape against the calendar

Put the link acquisition dates next to the publishing history and look for the seam.

Forty articles across three years and three hundred referring domains inside a single month is not a site that earned them. Somebody ran a campaign. Plenty of sites live through that, and it still changes what you are buying.

Check when the burst happened. A spike four years ago that the site has since survived is a different object from one from last quarter that has not been tested by anything yet.

The direction the tools do not open on

Now the part I care about, which almost no buyer runs.

Crawl the site you are considering and count what it links out to. Pull every outbound link with its rel attribute, group by destination domain, sort by count.

I ran that on my own portfolio and got 15,321 outbound dofollow links against 16 inbound. I checked three times, assuming the crawler was broken. It was not, and it had been that way for years on a property I sell placements from.

The volume never comes from the articles. It comes from templates. A footer credit, an author profile field that renders a followable URL, a sidebar widget, a plugin that started stamping a link on every page after an update nobody read.

The arithmetic is what makes it unintuitive. Two thousand pages and one footer link is two thousand outbound dofollow links from a decision that took thirty seconds. Four more of those over the years and you are at ten thousand, from five choices nobody wrote down.

That sort by count is where the answer falls out. Editorial links spread thin, one or two per destination. A leak shows as one domain appearing eight hundred times.

Then check the destinations resolve, and watch the ones that redirect. A domain that lapsed and got bought at auction still returns a healthy status code and passes any check that only counts failures. Your new pages would be recommending whatever the buyer built there.

Why outbound reads truer than inbound

Here is the position I will defend.

Anyone with a budget can produce a decent looking inbound profile. A couple of thousand dollars and a week of emails does it on a site nobody has ever read, and much of this market exists to make that possible.

Outbound cannot be bought, so it accumulates instead. It is the residue of every decision the previous operator made and never revisited, and it does not get cleaned up before a sale because most sellers have no idea it is visible.

A tidy outbound profile means somebody was paying attention for years. A site handing out thousands of unintended dofollow links through a widget means those years went unsupervised, and everything else that went unsupervised is about to become your problem.

You inherit the page, not the agreement

The domain transfers. The arrangements do not.

If the previous owner sold a twelve month placement with a named slot on a specific page, that deal lives in an email thread you will never see. Rewrite the page in month one because it reads thin and you have broken terms you never agreed to.

Swaps decay faster. A swap has no invoice and nothing to enforce, and it runs on two people knowing each other, which is why so many never complete even while both sides stay friendly. At handover that social pressure vanishes in one step, and every swap the last owner had becomes a link on your page pointing at somebody who owes you nothing.

So ask for the list before signing. What is on this site for other people, who paid, what was promised, when does it end. If the seller cannot answer, price the uncertainty in and expect to learn it by email.

Change almost nothing for a quarter

The urge on day one is to clean house. Cull the thin pages, strip the outbound links, file a disavow against everything with an ugly score.

Don’t. You have no baseline. Change eight things in the first fortnight, watch the site move, and you have learned nothing about which of the eight did it. Disavow files are their own subject and mostly a mistake, and a brand new owner is the worst placed person alive to judge which links a site earned honestly.

The exception is the template level outbound stuff, because none of it involves judgement. A plugin footer linking out on every page was never an editorial choice by anybody. Remove it, note the date, and leave anything with a human decision behind it alone until you can name a reason.

What I have not done

I have audited link profiles on sites I already own, and I have watched sites carrying my placements change hands from the other side of the transaction. I have not bought a site at the top of this market and run this list against a serious asking price.

So the outbound argument is the part I would stand behind under pressure, because that is where my own numbers came from. The rest is what I would do with my own money.

My outbound ledger only starts in July of this year. Everything before that is reconstructed from email searches, which means it is incomplete, and if I sold one of those sites tomorrow the buyer would inherit exactly the problem this article is about. The questions I send a seller before money moves are here.