Understanding Link Reputation

There was an interesting comment made in a private discussion list I’m a member of related to paid links. Someone I have great respect for gave an example of a page that had several links on it, one of which was paid for, the others which were not paid for. Same site, one paid, the […]

Chat with SearchBot


There was an interesting comment made in a private discussion list I’m a member of related to paid links. Someone I have great respect for gave an example of a page that had several links on it, one of which was paid for, the others which were not paid for. Same site, one paid, the rest purely editorial. The paid link was placed on the site on purpose, in such a way that there was not anything about the links that indicated which one of the links was paid for. The source code gave no clues. No sponsor box. No “Buy a link here” text or graphical evidence. The transaction was made privately between buyer and seller, and nobody else knew about it. Bottom line? The paid link on that page was 100% indistinguishable from the editorial links on that page.

The point the writer was making was this was a case where a paid link would in fact fool the search engines, because they’d have no way to distinguish paid from not paid. The person paying for the link would receive algorithmic benefit for a link they should not receive any for.

At a micro, page specific level, yes, the above scenario is true. On this one page, no algorithm could identify the paid link.

But wait a moment before you start having secret meetings in dark alleys to arrange a similar tactic.


The flaw in the logic

Search engines do crawl the web a page at a time, and the page above–if looked at all by itself– would seem to be a way to game the system. But search engines don’t necessarily base their decisions on a single page. Engines have a huge history of link data they’ve analyzed. Links I pursued and received for my sites or my clients ten years ago are still around.

You can learn a lot about a site just by looking at its previous collection of inbound links. Has that site shown a tendency to participate in tactics that engines know to be spam? Has that site paid for links in other ways that engines have found? Does a site show a history of having the same content show up on a new domain every year? Has a site had a recent surge in 301 redirects from other domains? In short, your site has a reputation based on your linking history that will help the engines determine just how much they can trust you today. So what if you have a undetectable link on one page somewhere? If historically it’s obvious you’ve peppered blog comments with links, added your site to 700 free-for-all pages, and syndicated articles like a human copy machine, then guess what? It is pretty easy for the engines to spot such behavior.

I am not saying the engines do this perfectly, or at all. But I have some pretty good data that says they do. I’ll cover this in far more depth than is appropriate for this forum, but let me close with an example.

Two fellows who don’t know each other are dressed identically in $1,000 Armani suits, standing next to each other on a corner. At first glance, they appear the same. Handsome. Hard working. Successful. However, one of them stole his suit, and is currently wanted by police for other offenses. He has a rap sheet 20 pages long. The other fellow is hasn’t missed a day of work in 10 years, has never been arrested, and got his Armani by saving for a year.

From looks alone at that single moment, nobody knows who is trustworthy, and who isn’t.

Web sites and their links are the same way. They might look the same on a page by page basis, but collectively, like a police record, they paint a picture of character and intent that can lead to rightful suspicion. So forget about the one perfect link that may fool the engines. If you’ve been up to no good in the past, and the evidence (in the form of links) still remain, odds are the engines already know. Engines crawl links, and make decisions about them. You may be smarter than them on one web page, but they are smarter than you on a billion others.

Eric Ward has been in the link building and content publicity game since 1994, providing services ranking from linking strategy to a monthly private newsletters on linking for subscribers, The Ward Report. The Link Week column appears on Mondays at Search Engine Land.


Opinions expressed in this article are those of the guest author and not necessarily Search Engine Land. Staff authors are listed here.


About the author

Eric Ward
Contributor
Eric Ward founded the Web's first services for announcing, linking, and building buzz for Web sites, back in 1994. Ward is best known as the person behind the linking campaigns for Amazon.com Books, Weather.com, The Link Exchange, Rodney Dangerfield (Rodney.com), the AMA, and PBS.org. His services won the 1995 Award for Internet Marketing Excellence, and he was selected as one of the Web's 100 most influential people by Websight magazine. In 2009 Eric was one of 25 people profiled in the book Online Marketing Heroes. Eric has spoken at over 100 industry conferences and now publishes LinkMoses Private, a subscription based link opportunity and strategy service. Eric has written linking strategy and advice columns for SearchEngineLand, MarketingProfs, ClickZ, Search Marketing Standard, SearchEngineGuide, Web Marketing Today, and Ad Age magazine. Learn more about Eric and his content publicity and link building services at http://www.ericward.com

Get the must-read newsletter for search marketers.