24

Does Google PageRank Affect Google SERP Rank?

With many marketers placing more weight on the SEO aspect of Internet marketing, I’ve decided it would be nice to do some SEO case studies. I am working on building a niche site for a follow-along SEO study as well.

For this study, I looked at the results of various keywords and analyzed the Google PageRank (PR) of each result, looking to find any correlation between where a site ranks in Google and the PageRank the site has.

Analyzing Google SERP Rank vs. Google PageRank

Many marketers explain that building a long-term income stream is more plausible with SEO than PPC, PPV, or any other paid traffic source. While the jury is still out on that one, building authority (or seemingly authority) sites usually have a lot of potential to hold high SERP (search engine results page) ranks for a long period of time when done correctly.

In my experience, the number of people focusing on and discussing PageRank has been trending down for the past 2 years, and this study may hold a little information as to why people are less worried about what their site PR is.

The Study

I looked at 20 different keywords, broad search results ranging from 5 million 3.5 billion.

I analyzed 2 different groups: 10 keywords with less than 100 million results, and 10 keywords with more than 100 million results.

I analyzed the top 10 results for each keyword for a total of 200 pages.

I included WikiPedia, Amazon, Google News, and Google Images in the results if they did occupy a specific rank spot.

I did not include local results, although they do in fact occupy various SERP spots.

The Keywords

Here are the keywords and corresponding results. Thanks to @gcarivey and @layeredseo for their keyword suggestions.

Under 100M (Low Group):

  • identity management: 5M
  • coffee cup: 13M
  • leather boots: 15M
  • chandelier: 19M
  • mortgage rates: 25M
  • mountain bike: 30M
  • curtains: 38M
  • green tea: 80M
  • sports car: 83M
  • fireplace: 85M

Over 100M (High Group):

  • nutrition: 195M
  • weight loss: 210M
  • Justin Bieber: 215M
  • tea: 300M
  • coffee: 350M
  • quotes: 410M
  • computers: 450M
  • iPad: 540M
  • Charlie Sheen: 585M
  • TV: 3.5B

Overall PageRank range for both groups was from 0-9, each broad group having at least once site PR9.

PageRank Findings

  • Average PR High Group: 5.2
  • Average PR Low Group: 3.8
  • Average WikiPedia PR: 5.9
  • 16% (17 pages from Low Group, 15 pages from High Group) were PR0
  • 22% of PR0 were Google News pages
  • 13% of PR0 were Google Images

Additional Findings

  • 75% of keywords (15/20) had a WikiPedia result
  • 4% of all results were YouTube
  • 9% of all results were either YouTube, Google News, or Google Images
  • 60% of results for ‘Charlie Sheen’ were PR0
  • 40% of results for ‘Justin Bieber’ were YouTube
  • Google News listing appeared most at SERP rank 4 (three times), and next at SERP rank 1 (two times)
  • WikiPedia appeared most at SERP rank 1 (four times), and next at SERP rank 2 and 6 (three times in each)

A Graphical Analysis

My assumption, and the assumption of many other marketers, is that a direct correlation exists between SERP rank and PageRank. In a perfect world, that would look something like this:

However, as with anything, especially in marketing, there is no “perfect world” and it is usually not safe to make assumptions.

Anyways, here are graphs of each keyword.

Low Group

High Group

When looking visually, it’s very easy to tell that not a single keyword analyzed, granted there were only 20, followed the ‘assumed’ relationship between PageRank and SERP rank.

Here’s the averages:

As you can see, the highest PageRank was actually in the top SERP spot.

However, based on the overall averages, it appears as though PageRank is not a clear-cut factor for SERP rank, although again, I only analyzed 20 results.

My Analysis

This study lead me to 6 conclusions.

  1. WikiPedia appears pretty damn often. I could be wrong here, but I thought Google penalized content farms with the last algorithm update. If you ask me, WikiPedia is a content farm of epic proportions. What’s the deal here?
  2. Analyzing 20 keywords may not be enough for this study. I realize it isn’t statistically significant anyways, but I did try to look at keywords from a variety of niches/industries, and used a wide variety of competition as well. Either way, I may have to look at this in more depth.
  3. PageRank may just be a good indicator of overall ranking ability. While the results were not staggering in favor of “a high PR guarantees a high SERP rank,” most sites with a higher PR also had more links, a longer domain age, and consequently, higher rankings.
  4. Keywords with more results generally have listings with higher PageRank. With the exception of Charlie Sheen (average PageRank of 2), no keyword in the ‘High Group’ had an average PR lower than 4.
  5. Google loves Justin Bieber.
  6. Niches with high profit potential and low competition still exist.

Although I did not take into account any other factors for SERP ranking, I feel the study was a success.

What’s Next?

First of all, I would love any comments about this study. If I receive good feedback, I’ll continue looking for similar factors to analyze whether it be with SEO, PPV, Facebook, POF, or whatever. If you liked it, simply leave me a comment, re-tweet it, or ‘Like’ it using the little widget on the left.

I truly appreciate any feedback or support YOU can give me.

Also, I am getting ready to launch my follow-along site. It’s a small niche site and I’m just going to try to rank it well for my primary keyword (originally).

Have a great week and weekend!

{ 24 comments }

Calgary Home Business March 16, 2011 at 12:00 pm

This is a GREAT post Jeremy.

My experience has shown that it’s just as easy for a PR1 to rank for a keyword as it is for a PR5 or PR6. However, I’ve also noticed that the higher PR sites ranking for competitive terms are doing so because of the link juice behind their position.

In my opinion, PageRank is a great indicator of website QUALITY and AUTHORITY, not necessarily of ranking capability.

Jeremy March 16, 2011 at 12:48 pm

Thanks for the comment man, and after doing the study, I have to say I agree.

I plan to evaluate ranking in more depth too, as this study only looked at one element.

Ed from htmlpress.net March 16, 2011 at 6:44 pm

re. Wikipedia : Google did penalize content farms – low quality content farms that is.

Jeremy March 16, 2011 at 9:18 pm

Wikipedia is considered to be high quality content?

Laurence 'GreenReaper' Parry August 1, 2011 at 2:23 am

It’s usually a lot better than the content farms.

It’s not pushing ads on you, and tries not to be biased. Readers appreciate that.

Jeremy August 1, 2011 at 7:07 pm

I agree, but Wikipedia is the definition of a “content farm” if you ask me.

Affiliate Paying March 17, 2011 at 6:42 am

“Google loves Justin Bieber.” – confirmed.

Jeremy March 17, 2011 at 8:51 am

Hahaha! I figured with 60% of his results that they do. I bet Charlie Sheen is next!

Carl Woolston March 17, 2011 at 9:18 am

Great research and very thought provoking. It would be very interesting to lay across the url names of those that didn’t correlate to page rank. In other words, take the lower page rank that made serp and my guess is they have the name of the search term in their url.

Jeremy March 17, 2011 at 9:31 am

Carl,

That’s a very interesting theory and I was actually just discussing a similar observation earlier with someone. I’ll keep that in mind and potentially evaluate another group of words/ranks to see if that is actually the case.

Maybe then I/we can delve deeper into why a lower PR site (other than YouTube/Images/News) would rank highly for a certain term.

I also plan to dig deeper into the backlinks/domain age and see if that factors in as well.

I appreciate your comment, thank you for stopping by! :)

Codrut @ Blog Commenting SEO March 27, 2011 at 4:46 am

That’s a good case study, I’d recommend you focus on long-tail keywords and put them into quotes for exact matches, this way we’ll have a big picture of what counts the most: page rank, back-links, domain titles, etc.

Jeremy March 28, 2011 at 12:57 am

Thanks Codrut, good idea!

Phil March 18, 2011 at 1:48 pm

Wiki wasnt hit hard as they dont monetize via adsense.

Google wanted to hit all the big content farms with adsense, therefore slashing the costs they need to pay out monthly to anyone monetizing from adsense.

Also the fact that many SEO’ers will link to an authority site (as this helps your sites credibility), usually its the wiki page related to their topic. Since Wiki gets thousands of links from unique IPs, the power in numbers comes into play as well.

Jeremy March 18, 2011 at 3:22 pm

Thanks for the comment Phil.

I understand that as well, but decreasing the number of sites serving AdSense ads also decreases their revenue, right?

Ricky Ahuja March 26, 2011 at 11:14 am

Quite informative and my daughter loves JB also – go figure!

PR is one factor in overall keyword ranking – not the be all, end all, but a good indicator regardless.

Jeremy March 27, 2011 at 1:29 am

That’s what I’ve come to realize as well. It seems that generally speaking, as one particular factor increases, others do as well.

Jared Carrizales March 28, 2011 at 2:59 pm

The Panda algo update is stated as being an attack on content farms, but I think in reality it was targeted at the spammers and keyword stuffers. Some argue (including myself) that this was an update to attack poor content attribution. Meaning google is working on/has somehow found a way to track the origin of a piece of content to its original publication.

So if this is in fact the case, any honest SEO shouldn’t have to worry about this update at all when writing or submitting content for a client. I think it’s hilarious that some of these people got hit hard on Panda. 99% of them are probably scammy, spammy, scrappers anyway. This is a great update I believe. My Squidoo, ezinearticles, articleslash, etc, haven’t been affected at all weeks after the fact.

Didn’t mean to go off on a tangent there, lol..

As far as PR goes, it’s just link popularity guys. LINK POPULARITY. Link popularity (PR) can be extremely vulnerable to manipulation. I can’t tell you how many “Sell Your Gold” sites I’ve seen that have PR4+ and rank worse than 3 million Alexa.

Just my 2 cents. Enjoyed the post

Jeremy March 28, 2011 at 5:42 pm

I think it’s more than just link popularity, as I’ve seen sites with 100k+ links that are PR0 still, and sites with far fewer links that are PR3-5. It’s a matter of mixing the two in my opinion.

Thanks for the comment Jared.

PPC Ian April 5, 2011 at 7:37 pm

Great post, Jeremy! What a great analysis. I have experienced results in my personal portfolio of sites that back up this study. Basically, PR can make the difference on high competition keywords. However, when you’re looking at the longer tail, I have sites with low PR that do better than my high PR ones. Also, I’ve seen the PR of my sites go up and down over time with no affect on traffic. These days, I think PR is mostly good as: a measure of how well you’re doing (generally), a factor in selling ad space (advertisers are impressed by high PR), and a factor in ranking on very competitive terms.

Jeremy April 5, 2011 at 8:15 pm

Exactly Ian, it’s a very big selling point in all of the advertising I have seen.

Thanks for the comment!

Fazal @ Make Money Online April 10, 2011 at 3:39 pm

thanks that is a nice analysics. As for my, I don’t think PR matters much, I know it’s an important ranking but it’s with better serp’s that your pr and alexa ranking will increase.

Jeremy April 11, 2011 at 3:31 pm

Well, good point but it’s not a completely linear relationship between SERP rank and PR. Thanks for the comment!

toto December 8, 2011 at 11:51 pm

very good analyzing. I dont know about google SERP rank, but you explaining very clear from description and graph from your post.
Thank you very much

Jeremy December 9, 2011 at 6:06 pm

Thanks for the comment toto – Do you have additional input on SERP rank factors?