February 22 2010
Diff Between Link Farm and Blog Farm
What is the difference between a Link Farm and a Blog Farm? Well, if you browse the SEO Theory SEO Glossary you find entries for both “link farms” and “blog farms”.
Blog Farm – Noun phrase. A group of blogs operated by a single person or group that are populated by software, usually RSS-feed scraping scripts. Used for link building, blog farms are created by special software that installs popular blogging software on multiple domains and hosting accounts. Sometimes confused with links farms (q.v.).
Link Farm – Noun phrase. Any group of Web sites where every member site in the group links to every other member site in the group.
Keeping in mind that there is no central authority for deciding which phrases actually mean what, we should accept that people are going to disagree on their definitions from time to time. In particular, academics who publish research on detecting and fighting Web spam often (in my experience, almost always) use terms that were developed by the Web spam community inappropriately. The academic community’s terminology is often adopted by search engineers at the major Websearch services.
Except for the fact that their blog posts may cause the rest of to feel puzzled, the search services are generally perceived as both authoritative and knowledgeable in these matters. For example, Bing search engineers sometimes share their views on Web spam. These posts are quite enlightening but as you can see from the comments I could not refrain from offering what I hope was a gently worded clarification in terminology.
My purpose is to ensure that people see both sides of the discussion, not to sway search engineers to use my terminology. They are dealing with abstractions most people never see. But the confusion over terminology is easily found around the Web. For example, a recent discussion at HighRankings concerned link wheels. Again, I offered a counter-example of how the phrase can be defined (and, in my opinion, how it is properly defined).
I don’t coin most of these expressions. I learn them from other people, watching how they use the terms. Take “link farms”, for example. The earliest link farm I know about would be Brett Tabke’s Buddy Links program, which he put together after I shared my findings on a forum about how I was able to improve my Inktomi stability by using crawl pages on different domains. Link farms became all the rage for several years and they morphed into different types of content.
Some people today ask what is the difference between a link farm and a directory, for example. Some link farms were designed to look and function just like Web directories. There was one popular program that would crawl the Web looking for Websites with reciprocal link pages, add those sites to a custom “directory”, and then send an email to the sites asking for a link back. Clever, right? It eventually led many sites down the golden path to search invisibility through bans and penalties.
Some — but not all — reciprocal link management programs were also link farms. Reciprocal link management services followed several different paths but the earliest one I know of (no longer in existence) grew out of the link farm concept. As the years passed reciprocal linking services became more selective and, according to some of the service providers, less interested in building links for SEO and more interested in building links for traffic and branding.
Search engineers in academia and industry hold an annual conference called AIR Web (AIR = Adversarial Information Retrieval). I am sure they feel they derive a great deal of value from the conferences so I don’t mean to belittle that perceived value.
I have no real idea of what goes on inside a search engineering organization. But I have noticed from the published papers that the research usually deals with spam techniques that are 2-3 years old, if not older. The apparent lag time between the widespread adoption of the latest Web spam techniques and the identification of those techniques by researchers is somewhat concerning to me, but then just because state-of-the-art spammers have moved on from 2007’s tactics doesn’t mean they are irrelevant to today’s Web.
What I find more troubling is that the AIRWeb papers (and other published research) often use definitions of Web spam tactics that are radically different from what the actual spammers use. That doesn’t mean these folks don’t know what they are talking about but it does imply they don’t know what the spammers are talking about — and vice versa. In other words, spam by any other name would still smell as foul (may the Shakespeareans in the audience forgive me for that). Web spam, however, seems to be staying ahead of search technology in part due to the differences in dialect between spammers and search engineers.
This plurality of technical definitions for the same concepts we all run into is very confusing, so it’s no wonder people often “discover” new ideas that are really old ideas. Still, one must also be careful not to judge too quickly.
For example, suppose what the Bing engineers identified as “link farms” in the post I commented on really are NOT blog farms (as I speculate)? The article did not provide enough information for me to be sure of what they were talking about — sure, that is, in terms with which I am familiar and comfortable. The distinctions between artificial Website paradigms blur very, very easily.
For example, you could easily set up a link farm using blog farm software, wherein all the blogs link to each other. You have both a link farm and a blog farm, but neither is the other — conceptually the blog farm should still be distinct (in your abstraction of definitive terminology) from the link farm, even though the instance of both is the same group of Websites.
Only someone as esoteric as I would care about that.
You might ask if the so-called “SEO friendly” Web directories that I have ranted against for years are link farms. Some of them are like link farms in that they required links back from listed sites, but they were still not link farms because they didn’t meet the technical definition of a link farm. That is, the listed sites don’t all link to each other. They just link to the faux Web directories.
People sometimes call these sites Spam Directories but I haven’t really see enough general usage for any one expression to feel justified in including such a term in the SEO Theory glossary. Dictionary editors sometimes wait several years before adding new words or expressions, to see if they really fit into general usage. I suspect we’ll never be able to document a widely used expression for “SEO friendly” spam directories (other than SEO friendly spam directory, but that is a pejorative term used by critics of the sites, not users of the sites).
The diff between link farm and blog farm expression arises from someone’s curiosity about link farms and blog farms. It is, to me, an indication of the widespread interest in Web spam. I am sure there are new faces appearing in the Web spam community all the time. I usually call those folks script kiddies because they often start out using the older tactics (and software) that hard core, experienced black hat spammers have moved on from or downgraded in their strategic arsenals of spam techniques.
This is a hard topic to write for because of all the different uses for common terms, as well as because of the plurality in terminology for common techniques and Web structures. I don’t expect academia and search engineers to suddenly change their vocabularies just to satisfy my ego but if you’re going to study Web spam then you should at the very least be aware that there are different dialects in the discussion.
Written by Michael Martinez




