Filthy Linking Nostalgia

filthy-richIn Bryan Eisenberg‘s ClickZ column today, he writes about being “Data Rich, Optimization Poor,” and chooses to link to a paper I wrote seven years ago: “Filthy Linking Rich!” Ah, the nostalgia.

Funny thing is, what started out as a blog post turned into an essay, which ended up as a 10-page document that has been downloaded tens of thousands of times since 2004. It’s interesting that, even after seven years, I still get regular feedback, not just from SEO types who have read it, but also students of information retrieval, computer science, and librarians. In fact, and this is true, I was in a restroom in Shanghai, China earlier this year when, amazingly, this guy blurted out “you’re the filthy linking rich guy, yes?”

Perhaps the question I get asked most about it is, “why did you write it?”

As a marketer and not a physicist involved in network theory or a scientist involved in information retrieval, it seems to have puzzled quite a few people. But the reason I wrote it starts at the end of the paper, actually.

I had been doing a lot of reading about hyperlink-based algorithms, in particular, trying to get a better idea of how Google worked. And, like many others, had come to the conclusion that Google’s approach to ordering the web and ranking it was not as democratic as they had originally proposed. In fact, I had come to the conclusion that there was a distinct disadvantage to certain web properties when linkage data was being used as the main ranking criteria.

It was at that time I read a paper written by a young researcher by the name of Junghoo Cho, now associate professor at the University of California. I had read his thesis on crawling the web and knew of his work.

Cho’s paper, “The Impact Of Search Engines On Page Popularity,” touched on the “rich get richer” concept I was thinking about. And I was already aware of this from the varied types of clients I was working with at the time. The larger more popular brands easily attracted many more links over a given period than smaller less known brands and products.

What’s more, the websites that had most links ranked more highly and were therefore discovered more often. And the more often they were discovered, the more links they acquired. Basically, they became “Filthy Linking Rich” by comparison to other web sites.

However, in order to explain this concept I needed to also explain a little more about network theory and the connectivity, or graph structure, of the web. Fortunately, I had been devouring the work of two brilliant minds in the fields of computer science and network theory: Jon Kleinberg and Duncan Watts, who I’ve actually come to know personally all these years later.

I even referenced the true source of the phrase “six degrees of separation” in the paper. Interestingly, as it had been brought to my attention again this week, I actually read it again. More interestingly, even with its quirky writing style, the topic is covered quite concisely and the argument of the rich getting richer (in link popularity terms, as it was quaintly known back in the day) still stands up today.

Related reading

The middle ground for single keyword ad groups (SKAGs)
Five ways blockchain will impact search marketing
New visual search innovations tap human emotions and biological buying triggers