I was impressed with this guide and think it's an important document. Answering a lot of Q&A, I often read gross misconceptions about how PageRank works. This guide puts many of those misunderstandings to rest. It is also probably our most diagram-heavy guide; the graphics really add to the quality of the article and help explain some of its complexities. Even if you know/understand very little about how real PageRank works (not the green bar in your toolbar which lies, cheats and laughs at you), the guide is both easy to read and to comprehend.
After reading this, there is really no excuse for not having a good understanding of PageRank and how to sculpt it for your advantage.
PRO members have access to the guide already, but if you'd like to purchase it separately, it's available in the articles section of our SEOmoz store. The individual article's page is here.
Happy reading: the 10-question quiz will be administered in a week from today :)
Outstanding. IMHO, the most valuable article for me in the SEOmoz Pro archive by miles.
Why? As a SEO, the question, "What is Page Rank?" is one I regularly encounter at work and leisure.
This article completely craps over the internal doc I use at work and pretty much any other explanation I've seen online. It's extremely high utility value as I can not only use it for introducing newbies to PR but it saves me time in explaining link architecture concepts to my search marketers and developers.
Darren, I intended to 'borrow' some of the images for my next internal SEO roundtable but I think I'm just going to use your entire guide. The before and after stuff is pure gold.
edit: And yes Mr Slatten, you deserve your title!
The before and after stuff is pure gold. Thanks, shor. I really struggled to keep my OCD/perfectionism in check when I was making the diagrams. If you can believe this... the spheres are all drawn to scale. In other words, I calculated the diameter of each circle, based on the PageRank value it was representing. So in those Before/After pictures, the spheres change in size relative to the change in PageRank. =)
Darren,
This is appears to be very solid, well-illustrated document.
It challenges some things that I have read from other sources. I was under the assumption that Google used rel-nofollow links to find content but didn't pass value to them (I *believe* I have seen newer pages - such as a privacy policy - with strictly nofollow internal links and no known external links show up in the Google results). And Rand has illustrated how pages lose a fraction of their PR when they link out to other pages - something that this guide disputes more than once.
I am definitely going to print this out, get some coffee and try to make clearer sense of all of it. Thanks for your in-depth research and efforts!
Brett - If I previously indicated that a page loses its own PageRank when linking out, I apologize - that is certainly not the case. It is true that when linking out with multiple links, you give up the opportunity to pass more PageRank to each page you link to, so perhaps I simply communicated poorly.
I think the other really good illustration on the blog regarding the flow of link juice is this post from Si - https://www.seomoz.org/blog/how-pagerank-works-why-the-original-pr-formula-may-be-flawed
Thanks for that explanation. I had the same impression Brett did, that we'd lose the PR. So many more things make more sense now. It brings a lot of other posts in a different light to me.
I am going to have to re-read the whole SeoMoz vault. Anyone want to write a guide to that? ;)
Thanks!
Brett, Thanks for the feedback. I'd be happy to address your concerns. =) 1st concern: My guide says Google does NOT use nofollowed links for discovery, but you have heard otherwise. My guide merely echoes the claim that has been reiterated several times by Matt Cutts and other Googlers. Here is an example, taken from an interview of Matt Cutts by Stephen Spencer: At least for Google, we have taken a very clear stance that [nofollowed] links are not even used for discovery; they are not used for PageRank; they are not used for anchor text in any way. Anybody can go and do various experiments to verify that. Even with that being said, I still can't figure out why Google Webmaster Tools reports inbound links that are nofollowed. For example, GWT reports a link from Matt Cutts' blog to www.seomofo.com, even though it was a nofollowed link from a comment I made. Personally, I try not to rely on the nofollow attribute 100%, which is why I mentioned the idea of "natural linking" and covered so many alternatives in the guide. Bottom line: there are a lot of conflicting theories out there about PageRank, but the theories in this guide are supported by published material from Google itself. 2nd concern: Rand has illustrated how pages lose a fraction of their PR when they link out to other pages - something that this guide disputes more than once. Right... well if Rand knows so much, why didn't HE write the f#$%ing guide? ...kidding! =) I don't know what you're referring to that Rand has said, but I will try my best to explain why we are BOTH right. In other words... my guide doesn't actually dispute the fact that a page's PageRank may drop if it links to other pages. There is a subtle difference here between a page "losing PageRank" and a page "receiving less PageRank." The best way to understand this difference is to think of a single-page website vs a multi-page website's home page. If you take your single-page website and link out to some random page, your PageRank will not drop, because you are not giving away PageRank from your own stash--you're giving away your distributable PageRank. On the other hand, if you take your home page from the multi-page site and link out to some random page, your PageRank WILL drop, because the other pages on your site now get slightly less of your home page's distributable PageRank, and therefore... they have slightly less to distribute back to your home page. So in this case, your home page didn't give away a piece of its own PageRank stash--rather, it ended up receiving less PageRank from the other pages on the site (who are now sharing with that random page). Now... be honest... did I even come CLOSE to answering your questions... or did I just make you even more confused? =P
The point you make in the last part of that comment is one of the good things the guide covers. I actually had a discussion about that on Twitter this morning; it's such a simple concept in theory and yet something that people fail to get their heads around.
Luckily... I have a ginormous head that wraps around virtually anything.
No, you gave a great answer to my questions. Thanks.
Sterling job Darren. It reads really well and the examples of logarithmic scales show PR in a new light to me at least. I'm only half way through at the moment, but so far so good!
Just one question... at the bottom of the diagram showing page A and page B (Accumulating PageRank vs Distributing PageRank) you state that if both page A and B link to each other then both will have PR=1.0
I don't quite understand this as wouldn't both either have a PR=0.15 or some cyclic PR reinforcement? How come they'd both have PR=1.0?
Edit: It helps if you read the next section on Convergence of PageRank ! lol. Very thorough Darren :)
Lindop
Good work. I like the clarity and especially the diagram (and yes, I appreciated the fact that the circles changed size). I tend to prefer to think of a rebased PR - i.e. rather than being potentially infinite, everyone's PR is scaled by that of the highest PR - so it becomes a number between 0 and 1, but that's just me. I suspect you can safely ignore me :)
I agree - though I don't think it makes any difference in application. Which I guess is true of a lot of maths......
Darren You really and I mean really nailed it.
Of course you didn't need me to say that and I am sorry to all the mozers that are going to get a notice that someone commented here and just said way to go, But Way To Go.
Most of my clients are small to medium size business that need all the help I can give to grasp what this is all about as quickly as possible. I am Always drawing pictures and arrows and trying to relate it to something they can understand without sounding condescending. This really is a Huge document.
I will "drive it like I stole it" to pass on the golden egg of wisdom you have put so much effort into. Really Top Notch. This is really why SEOmoz rules. phenomenal content.
If you are not a Pro member... sucks to be you!
So the World's Greatest SEO is at it again huh? ;)
Looking forward to reading this. Darren is the world's greatest SEO, after all.
Nice guide filled with great information. Thanks mozzers.
I was under the impression that linking out to other sites would make your page rank better. That brief little explanation Darren gave about sharing your homepage page rank with internal pages leaves me drooling with wanton desire to read the entire doc (a.k.a I was so wrong).
I've been trying to get my higher ups to pay for a membership ... more fuel for the fire here!
Huge work done and the whole day reading ahead... Loved the images :)
Nice, this one is worth picking up--
It took me a while, but I finally got around to reading this - really nice work, Darren. I'm not ashamed to admit that I learned a couple of things.
Wow. I left it pretty late to read this but it is one of the best SEO documents I've ever read. Thank you Darren, and SEOmoz.
Not sure if anyone is watching this post after all this time, but I have some specific questions about optimising anchor text in your internal links - eg if you have a header link to a page with not too great anchor text, and then another link to the same page later on in the copy or footer with A1 anchor text, how do you get Google to use the 2nd anchor and ignore the 1st? I've read that nofollowing the first link will automatically nofollow all links to that page...
Will ask this more coherently in Q&A...
Well you probably could've guessed that *I* still watch this post. =) I'll answer your question as best I can. I've dealt with similar situations many times. Here are some of the strategies I've used. 1.) You can remove entire blocks of the header code from every page, by placing the code in an external file and displaying it through an iframe. Here is an example from a site I'm building for a client who sells espresso machines. The row of brown buttons at the top, plus the search box, are actually an iframe. The HTML code for these links is not actually on the main page. 2.) Instead of adding the nofollow attribute to the first link (I've heard similar claims about this causing all links to be nofollowed), you can use JavaScript to make it "uncrawlable." Since Google is getting smarter about interpreting JavaScript code, you should define a simple link function in an external .js file (which you could disallow in your robots.txt file, if you wanted to). For example (here comes another shameless plug!), check out the breadcrumbs at the top of this product page for the Rancilio Silvia espresso machine. All of those links are crawlable HTML links, except for the "Home" link, which is generated by a JavaScript function that isn't defined anywhere in the page code. There's no way for search engines to know what the function does, unless they request the external .js file. This allows me to use the anchor text "Home"--which is easy for users to understand--without negatively affecting my SEO efforts. I hope that helps. Thanks for the feedback on my PageRank guide. =) -Darren
Nice.
This is why I never stop following this stuff.
Another golden nugget.
Correction: This is why you never stop following ME! P.S. Put your thumb where your mouth is. P.P.S. That means you should give my comment a thumbs-up if you liked it. It doesn't mean whatever you were thinking. You weirdo.
I am signing up for the twelve step program at Stalkers-R-Us.
It has not been a guilt free day.
Awesome. Who needs PRO Q&A's when you've got a direct line into the brain of an SEO mofo?
(The gratuitous link above is coz I wanted to test why your links above are not nofollowed - is it a perk you get for writing such a top guide for SEOmoz?)
The gratuitous link above is coz I wanted to test why your links above are not nofollowed - is it a perk you get for writing such a top guide for SEOmoz? That's a good question. I looked at all the other links I've left in comments, and they all have nofollow attributes, so I don't know why this thread is any different. Maybe you're right--maybe a dev flipped a switch or something because I contributed the PR guide. It only cost me about 200 hours of my life... what a bargain! Too bad this post doesn't have any PageRank.
Wow, I'd just like to say that that is without doubt, one of the most comprehensive and easy to understand in-depth PR explanations I've ever read. High production value as well, which is a nice touch. I'll definitely be using it to educate others.
Great work. Thanks guys.
Great guide. The images really do a spectacular job of showing how all the math works.
Holy amazing content batman. Great insight on what the REAL PR is and how to take advantage of it.
Just so I'm clear, when you're discussing PR for a singular system, are you talking about what Rand called Domain PR? If not, can you please describe the diff. between the two? thx.
Cheers,
@trontastic
Hey, Levi. Glad you like the guide. =) When I make references to a "system" of pages, I'm simply talking about a group of pages that link to each other and nothing else. This isn't something you're likely to find in Google's index, but it simplifies the examples I used. For instance, when I show the real PR values change from Before to After, those values are calculated as if those pages are the only pages in Google's index. Again, this was done to keep things simple. (The reason I used the word "system" is because I was imagining PageRank to be like energy. When you learn about thermodynamics, you often have to calculate the gains or losses of energy in an isolated system.) So I guess the total PageRank of a system would be more like what Rand referred to as Domain Juice, since it represents the sum of all individual PageRanks. However, keep in mind that I never really understood what Domain PageRank was supposed to be... and I eventually killed that entire thread with my overdose of nerd-speak. =)
Okie Dokie. Thank you for the explanation, good stuff. But now I have another one for you. If www.randomdomain.com is linking to your site with tracking code that encrypts the link, how does the PR flow? Here is an example: Link is on the page www.randomdomain.com/awesomelinks/goodcontent.html in the source for that site it shows a tag like this:
...a href="https://www.domain.com/ccode/tracker.php=2345">anchor txt ...
but when you click on it, the link jumps to your site - www.seomoz.org/blog/awesomepost/
Anyone?
If that "jump" is a 301 redirect, then theoretically, the PR would be forwarded to the final destination URL.
for some genius reason, I didn't even think to check to see if it is a 301. After checking I found out that the links in question are all 302s. LOVELY.
Thanks D.
You Critchlow boys (or boy if the rumours are true) and you're crazy theories...
I'm looking forward to reading this Darren - sounds like a blinder from the comments above.
Without being too modest... I'd say this guide contains the greatest explanation of PageRank in the known Universe.
I don't think that word means what you think it means ;) but you're probably right.
Nice one, thanks guys.
Very well written and useful guide. The diagrams were very helpful.
Kudos to Darren on a job well done.
Darren,
Well worth the wait and worth it's weight in gold.
This is a brilliant piece and if If I were the power that be at Google, I would ask Rand "What will it cost us to buy the rights for this document for open distribution?"
It's that good.
...worth it's weight in gold. Hey, wait a minute... what if you print a copy on paper made out of platinum?
Wow a PageRank optimization guide written by the world's greatest seo,i have to buy PRO membership instantly :D
I think i will need to earn PRO membership now. Happy Blogging to me n all.
You've come a long way from 'outlaw' to 'writes PRO material'.. Congrats, great material!
Woo Hoo! Pro memberships at work. This is exactly what I needed to educate the team on what's behind that green bar. Much more in depth than I had expected.
After reading all the comments, this guide is sure on TOP of my must have List.
Optimzation not spelled right in post title.
Heh. Fixed.
Can't wait to get the pro membership!! You guys RAWK!
Hi I would like to be a PRO member. I want to know if the pro member has access to every single articles ever released by SEOmoz or only the articles for that month/membership-time period?
If I have access to all the articles inthe archive (PDF?), how many articles are there in the archive? Sorry for the dumb questions, I am still new to SEOmoz.
Thanks!
Right now we have seven PRO articles and if you're a PRO member, you have unlimited access to all the PRO articles. They're not in PDF format, but are in .doc and OpenOffice formats for download :)
Thanks Mozers, I am working on a project right now designed to increase PR for a Website and was looking for better ways to sculpt the Site's PR. Guess will not sleep tonight.
The only problem is that PageRank doesnt move your website any high
Flip backwards over that railing?