All Entries Tagged With: "msn"

BrowseRank Would Discriminate Against Blogs

Chris McElroy made a plug for BrowseRank on yesterday’s SEO Service Provider blog. Like many people in the search business, I believe that search is ready for a major innovation, but I’m not altogether sure that BrowseRank is the answer. At least, not the complete answer.

In the early days, before the Web went commercial, web pages were ranked according to how many academics thought the page was important. It made sense to do it this way because the Internet was largely a research tool for universities and the military-industrial complex. Since the majority of users were academics, more weight was placed on what academics considered weighty or important.

The second wave of what search engines considered trustworthy came when Google founders Sergey Brin and Larry Page invented BackRub, a tool that analyzed backlinks and use that information to rank web pages. This was a huge innovation. It was based on the previous incarnation mentioned above, but expanded on that to include links from non-academic websites. Shortly after, the Web was Google-ized and that algorithmic innovation became the norm for most search engines. New innovations that have been tried since then have tried to counter the importance of backlinks, but those innovations – every one of them – have failed to catch on popularly.

The State Of Search Today
The problem today is that many webmasters have learned to manipulate search results through uncanny backlink practices. Plus, the more-than-10-year-old algorithmic interpretation of results has made the Web a bit of a wild west due to an unprecedented increase in the number of web pages to be ranked, and a huge volume of link data to analyze, as well as a wide variety of types of websites and intent with regard to user interaction. Google’s algorithms tend to favor older sites with a lot of backlinks. That’s a problem because many newer sites are worthy of trust and recognition, but there is something to be said for longevity. So where is the balance?

The answer is, there isn’t any. Blogs and other temporal information can achieve high rankings on a short-term basis, but to achieve long-term success you’ll need to optimize your blog completely, not just each individual post. Still, there is a huge difference between the nature of a blog and the nature of a static website. So too there are huge differences between the nature of social sites like Facebook and MySpace and sites like Amazon and eBay where users may just show up, make a purchase, and leave.

Why BrowseRank Will Hurt Blogs
On the surface, BrowseRank seems like a good idea. The problem is it will tilt the weight of trust and reliability to sites where users are encouraged to remain for a long time. That doesn’t include blogs.

It is a statistical conclusion that blog readers typically read one post then they are gone. The bounce rate for blogs is very high. Many social bookmarking sites have the same problem as many bookmarkers will show up just to submit a story then leave. That this wasn’t the original purpose for bookmarking sites is irrelevant. The point is, that’s what users do. Should the sites be penalized because users don’t stick around long enough to make them more credible and “trustworthy”?

This phenomenon, of course, wouldn’t apply to Facebook or MySpace since users of those sites tend to stick around longer and use the tools available – creating applications, making friends, approving friends requests, etc. But what about auction sites and consumer sites where users just show up and buy something then leave? With BrowseRank, those sites might penalized and consumer review sites could end up ranking higher than consumer purchase sites for the same search term. Consumer blog sites would fall to the bottom.

While PageRank has its problems, I can see that BrowseRank will also have its issues. Those issues include, but are not limited to, favoritism of one type of site over another, the ease of gaming the results, and lack of human analysis since algorithms will do most of the work. Those are the same issues we have now with PageRank, but the difference will be that the problems will tilt the balance of favor from one type of site to another. Instead of older sites being favored as with PageRank, sites with lower bounce rates would be favored, but a low bounce rate is not always a bad thing.

Is There A Middle Ground Between
PageRank And BrowseRank?
I favor a combination of backlink analysis with on-site user behavior analysis. I do not necessarily mean the length of time that users remain on a site. There are other factors that are important for judging user behavior. For instance, do users tend to click internal site links or site exit links such as AdSense and display ads? If an algorithm favored the former then that might kill all those Made For AdSense sites that showcase useless keyword-stuffed content. On the other hand, it would also kill legitimate sites where the owners did a poor job of optimizing the content to encourage users to stick around longer instead of clicking the sidebar AdSense ads. That might be a good way to encourage better content.

What it necessary, I think, is a way to analyze the intended nature of a site and give weight to factors that are important to that nature. For instance, what is important for a successful blog is completely different than what is important for a successful static website. Perhaps one could be judged by the number of backlinks while the other is judged by the length of time users remain on the site. But if that static website is a consumer site where users are likely to show up and buy something then leave then perhaps it would be judged by another set of criteria entirely. This is somewhat what Google already does. Since Google analyzes over 150 search factors for any website on any given day, there is always a chance that a particular site is judged by what it does successfully AND by what it does half-heartedly or not successfully at all. It is the aggregate of the algorithmic analysis that is important, not the specific criteria.

I think we can all give kudos to MSN for attempting to take search in a new direction. MSN is certainly in a better position to challenge Google’s dominance than a new startup. The problem is that BrowseRank, in it’s current form, is incomplete. MSN could be on the right track, but before they commit to BrowseRank, they’ll need to put more thought into the nature of websites and the purpose for interaction in the first place.