Wednesday, August 23, 2017

New Google PageRank Algorithm Debunked

In my recent article on SEJ about Google PageRank I suggested that a new formula could be in action. The article has gone big on Twitter and LinkedIn, with hundreds of shares. And from some comments, it looks like many people accepted it for a fact.

Now I hate to break it but I have to admit that I was most probably wrong. So I need to write this follow-up article to prevent yet another SEO myth from spreading. Let me explain what happened:
  • I saw an unreasonably low toolbar PageRank value for my blog after the last update;
  • I started looking around for possible reasons;
  • I noticed many people observed similar effects;
  • I found an article by Bill Slawski about Google’s Reasonable Surfer model, which seemed to explain my observations and other people’s rankings well;
  • I decided to share my findings for consideration and discussion.
     And the discussion followed! My article collected about 100 comments, and a few SEO experts came up to correct me. The problem is, too many people accepted the alleged news without doubt and spread it further, so now it is my responsibility to get things straight.
    So what was the real cause of poor ratings? 
    Special thanks to Donna Fontenot for providing the most likely explanation for the effect that I and other webmasters observed. The toolbar PageRank value is known to lag substantially even after an update. Matt Cutts mentioned it in his blog post about what is a Google update.

    Please note the difference between the toolbar PageRank (TBPR) and Google’s internal PR value used as one of their 200+ factors to rank pages. I should have made this distinction more clear in my original article, too. Toolbar PageRank is more or less useless today because of how rarely it gets updated, and how much it lags. You can safely ignore it. If you are good with your link building, you will notice that from improvements in your search rankings first, and toolbar PageRank value will catch up eventually.

    This devaluation of toolbar PageRank even makes quite a few people declare that “PageRank is dead”. This includes some industry experts. Be careful when listening to these proclamations. Those people know what they are talking about, but they are referring to toolbar PageRank value, not the Google’s internal formula. Internal PR is still a significant ranking factor, albeit only one of the hundreds.
    And what about the Reasonable Surfer model? 
    I based my previous explanation of poor TBPR ratings after the recent update on Google’s Reasonable Surfer patent. But the thing is, a patent does not equal implementation! Google has hundreds of patents. They even seem to have a weight loss related patent. This does not mean they are secretly crafting weight loss into their ranking algorithm.

    On one hand, some parts of the Reasonable Surfer model could have been tested by Google even before they filed the patent (and that was in 2004). On the other hand, some parts of it may still not be implemented, either because of engineering difficulties, or because of too much noise in the suggested signals. I don’t know the current status of this model’s implementation, and I’m afraid Google will not disclose it.

    I, for one, would welcome the change if they crafted this model directly into the PR and TBPR formula. That would make the link building efforts of those webmasters who are focused on TBPR more productive. It’s an easily observable single indicator, and it is often used for bragging and comparison, so it would be nice if it correlated more with the true quality of the page’s link profile.
    In any case, it is worth your time to learn about this model, because it has a good chance to influence your rankings one way or another, now or in future. Just don’t take this for a hard fact. Here is one more good article about this model: The reasonable surfer; makes for unreasonable thinkers.
    What practical conclusions you can make from this? 
    Number one conclusion is: do not trust everything you read. If someone publishes an article on a reputable site, that does not make him an expert. If someone is an expert, that does not mean he is always right. And even if someone is right, that does not mean the same applies to your situation. Be especially careful about trusting any anecdotal evidence. Double-check any information you plan to apply to your business or communicate to your customers.

    Next, while the Reasonable Surfer model may not be fully implemented yet, the work performed in that direction indicates that search engines are not happy with simple mechanical ratings, and want their rankings to match the human behavior closer. While for now you may still enjoy some results from link spam, rest assured that search engines will be fighting to make it obsolete. So do not make your business or your customers depend on cheap links entirely.

    One more reminder: don’t pay much if any attention to toolbar PageRank values. If you want to brag about your site, talk about unique visitors and conversion rate. If you want to find influential online partners, ask them about the same. That green bar just doesn’t mean too much today.
    And finally, if you write an article about a high-impact subject like PageRank, make all the necessary terminology distinctions very clearly, and consult real experts in the field before publishing, even if you have years of relevant business experience under your own belt.

    Credits Thanks to Barry Adams for raising an alarm about my previous article in a way that no-one could ignore. Thanks to David Harry for dissecting it at SEOBS and for reviewing this follow-up (Editor’s note: the link to SEOB post was added by Ann. Cheers, Dave!). Thanks to all the readers for your support and understanding!
    If you retweeted or otherwise shared my previous article, please share this one, too. Don’t let yet another SEO myth spread over the Internet. In fact, share it anyway, it will not hurt. Thank you!

    No comments:

    Post a Comment