In the digital marketing industry and especially within the SEO community, thought leaders and experts are constantly running tests and providing information on Google updates and factors that it may or may not be including as part of its ranking algorithm.
During a recent Google Q&A published by the Agency of Internet Marketing, industry thought leaders Ammon Johns, Eric Enge and Rand Fishkin had the opportunity to discuss the death or Toolbar PageRank, RankBrain and search quality metrics as ranking factors with Google Search Quality Senior Strategist Andrey Lipattsev.
I decided to distill it down and provide what I felt were the most valuable insights. In many cases, I’ve paraphrased much of the Q&A to make it succinct and to include what I felt was most meaningful.
The Death of Toolbar PageRank
Ammon: Why remove the PageRank Toolbar if PageRank is still an important part of Google’s ranking algorithm?
Andrey: Why do you think it was useful in the first place? It wasn’t so much removed as it died a natural death. It wasn’t bringing very much value internally and it wasn’t a valid benchmark for a site’s usefulness or how well it would rank. It was getting used as a currency by some SEOs that was really meaningless.
Eric: Removing Toolbar PageRank significantly increased the value of Moz’s Page and Domain Authority metrics. 🙂
Ammon: I used the toolbar to evaluate how well a site’s pages would get crawled and indexed.
Andrey: PageRank has no bearing on how often a site gets crawled or how quickly pages get indexed. I hope that improved metrics provided in Google’s Search Console reports will be what search marketers are looking at like clicks and impressions.
CTR and Pogo Sticking (SERP Bounce) as Ranking Factors
Eric: Can you provide insight into how click-through-rate can influence a page’s ranking and search quality?
Rand: In your opinion, why is it the case that most of the CTR experiments I’ve ran have resulted in a temporary ranking increase for a query mapped back to the page that’s getting clicked on? What’s happening here?
Andrey: My best guess is that the general interest resulting around a subject that you’re generating is creating the signals we’re looking for including mentions, links, social shares, etc. I suppose it throws us off for awhile until we are able to establish that none of that is relevant to the user intent.
Regarding click-through-rate as a ranking factor, we’d need to have broad enough and reliable enough data on bounce and CTR rates on the vast majority of websites everywhere in order to be able to make meaningful comparisons all of the time. It’s gameable and a tough challenge to make a very strong signal out of user engagement (CTR and SERP bounce).
Regardless of how many patents are filed, unless Google Search engineers come to meetings with strong data or statistical proof that suggest a change will improve search results, it’s not going to fly.
Ammon: We’ve heard that RankBrain is the 3rd most important signal contributing to Google search results. Would it be beneficial to know what the first two are?
Andrey: It’s content and links pointing to your site. There is no order. The way we can look at in a useful way is that Google is trying to get better at understanding natural language and understanding the meanings behind particular inputs (think voice search). You don’t generally don’t type a question the same way you would ask it. What RankBrain does is influence our understanding of a query.
Rand: Based on conversations I’ve had with Gary Illyes, I think the idea of RankBrain is that it can reorder or reweight ranking elements in order to product more relevant results.
Andrey: When you look at it like that, you’re prone to say, yeah it is a ranking factor. The funny thing is that it’s looking at not only what’s on the page, but how well it met the expectation of searchers. Did the results match the intent?
Eric: My understanding is that RankBrain doesn’t change the core ranking algorithm, but it might change the core ranking’s algorithm you choose to apply to the query, because your understanding of the query has changed.
Rand: Are you using machine learning in other areas of Google Search like spam, links and content?
Andrey: We are looking at using machine leaning across the board including images, web spam, etc. In web spam I can’t point to any huge success, but I’m confident we’ll keep trying to apply it.
Upcoming Mobile Update
Eric: Anything you can share about the upcoming mobile update?
Andrey: We are going to make mobile friendly pages more visible/prominent in search results. This should be released around mid April.
Wrapping It Up
All in all, this was a great discussion. BIG thanks to WebPromoExpert for hosting the Q&A.