Surprising SEO Case Study – What’s working in Google Now?

Video 10 - Posted on BHW

Before - Not Ranked
After - Deleted

Sorry dude but not a valuable comment. That's neither true, nor helpful to anyone here trying to make money.

Interesting video. I wonder if that would work out the same for regular websites.

I honestly doubt it because Youtube is such an authority domain but it's at least a loophole for now.

Backlinks are not all created equally. Some are better than others.

Knowing the difference between good and bad backlinks is what sets apart the average webmaster with a professional ranking company. If you don't know the difference you're blindly throwing darts hoping one will hit what you want. This also leads to knowing what you're targeting and being sure you can rank a keyword before putting money into it.

Exactly. It seems some people missed the point of the case study. Would you rather have 10 profile links or 10 links from CNN? Why? How do you know one works better than the other.

We just wanted to be sure what works.

That case study makes perfect logic. Whatever people used to call "social signals" were nothing but backlinks.
I believe I posted a thread somewhere, mentioning that Twitter became SEO unfriendly.

Just search for "site:t.co" on Google, and limit the results to the last month. Notice the trend? Now look at t.co/robots.txt.

Twitter social signals died together with its new robots.txt. Facebook has new nuances as well - likes are not gonna add you links. Same goes for Google's +1.

Social presence is good only for one thing - to create more social presence. It has nothing to do with SEO anymore.

Definitely explains our findings.

Nice Informative Video, But still as he said we cannot come to some conclusions on the base of that. Good to see PR Working like a charm. :cool:

I'd say while you can't make huge conclusions out of it and we know that, it definitely says contextual links are more valuable than other types of links.

Awesome share,although I think there are other things that should have been looked at, but none the less thanks for the effort!!

We're always making internal case studies on what works. Please share what other factors you think are worth looking at and we'll go ahead and test them for everybody here.

it's funny how SEO's only know WHAT to do and not WHY they're doing it... well they could say they know why they're doing it... because it works... but WHY does it work ?

Ever thought that instead of shooting links at websites in order to see WHAT effect they have on the rankings you people might be better of trying to understand WHY and HOW those links influence your rankings ?

I read all that BS at places like SEO journal. They like to have a big circle jerk of SEO theory. The point of real studies like this is that Google and Matt Cutts generally say one thing and do another and it's better to learn what actually works in the real world.

Sure I could get real high quality media links like I do for some of our clients but how does that help somebody here on BHW trying to run an affiliate site where that's not an option?

Is there a text version of this, unable to watch video till later this evening

Sorry no but the short version is that contextual links proved to be the big winner.

My videos are not even getting indexed, I have run a 500 article and 500 bookmarking lets see what happens.

Don't forget to follow good Panda practices and diversify your anchor text with a lot of generic stuff.



We're still open to suggestions on whatever else you guys would like to see us test next time FYI.
 
What are the safe panda practices? Would it be too cheesy to ask what type of onpage seo you did for your video?
 
What are the safe panda practices? Would it be too cheesy to ask what type of onpage seo you did for your video?

It's all cool.

For this case study, we just put a sentence or two description and tagged the video with our various keywords. All of the video descriptions used as tests were identical so it was major duplicate content. The tags were also identical. We also made sure the title contained our keywords.

Our keywords were in this format:


  • product name
  • product name review
  • product name reviews
  • product name scam

Normally we would also send retention views, like, subs, and favorites to these videos as well but for the purpose of getting an even test, we didn't in this case.
 
It's all cool.

For this case study, we just put a sentence or two description and tagged the video with our various keywords. All of the video descriptions used as tests were identical so it was major duplicate content. The tags were also identical. We also made sure the title contained our keywords.

Our keywords were in this format:


  • product name
  • product name review
  • product name reviews
  • product name scam

Normally we would also send retention views, like, subs, and favorites to these videos as well but for the purpose of getting an even test, we didn't in this case.

Cool study man! One question: did you do anything to help the links get indexed?
 
What are the safe panda practices? Would it be too cheesy to ask what type of onpage seo you did for your video?

G has a PDF document that lays out best site practices. The PDF shows how to do onpage SEO likemake a silo, tags, descriptions, etc. I found it last week randomly browsing around Webmaster Tools.
 
Like any good case study, this one points to the need for more research.

Thanks for this one. I hope there's more to come.
 
Wouldn't knowing more about the links you used be more important that anything you shared in the video?

I would think Authority would be something I tested, relevancy, and number of links...

Giving each link a value of 1... wouldn't you want each test to be the same end value of Links in order to know link juice?

Article and Press R - Most Links... Makes perfect sense - With an authority domain you could SPAM the ISHT out of it and possibly do nothing. If those link were never indexed. If you ran a manual scrapebox however and then indexed them -- you could actually gain a ton of ground.

PR hasn't been updated in over a year now almost - this number is meaningless. Authority is again a made up number from another SeO studier...

Why not run the successful test in combo towards the "safe" video and see what you get?

Why not try and rank an article on blogger or some other site the same way?

Please make useful test and don't lead the newbies astray... Now they are going to go away from some of the links that "didn't work" when if you ran do follow wiki's and got them indexed you could have ranked your video - done it before for my own test. And yes - there are do follow wiki's....

I think there are too many variables not disclosed to really gain anything from this test - Articles and Press Releases to rank videos on page 1 would be a starting conclusion -- but we don't know anything about the links or why it worked so it in itself is still like slipping and finding a 100 dollar bill then looking for more places to slip.

Not negative - just want everyone to understand a true test that would result in true information.
 
it's funny how SEO's only know WHAT to do and not WHY they're doing it... well they could say they know why they're doing it... because it works... but WHY does it work ?
That's interesting that you said that. I've always thought that because SEO is so transient, most people don't have time to learn why their methods work, so they just use those methods as a short-term strategy without completely understanding them.
 
Finally something useful shared. Yeah,short term seo is most less risky in terms of investment/roi wise.

It would ve been better if you had done PR and articles separately to see actually which one brings juice. :cool:
 
Last edited:
Here's a hint for you guys: +1s, tweets and likes don't work for SEO, but social shares (when somebody actually links to your post from Google+ or Facebook, or shares it on his/her own wall) actually work wonders.

Also, I see so many people tackling backlinks and Penguin, when in fact their rankings are plummeting because of Panda...yes, that's right, PANDA, not PENGUIN. I can only tell you that Panda is mostly driven by user search metrics (Google owns Chrome and has invested a lot of money into Firefox, A.K.A - they have the backend data on Firefox users as well) - hire a UI/UX designer make the user experience the best possible, so that the user metrics increase in your favor. There are some tools out there which you can use to manipulate these metrics.

Forget about backlinks, you only need around 10 of them which are relevant (hint, a private blog network consisting of expired domains with high DA, PA and Trust flow without a long record of being expired)- that's enough for the backlinks. Then focus on social, but not those crappy likes and +1s, but on real social link sharing on people's own walls. Finally, focus on good user experience to increase your user metrics stats, or manipulate it.

The way to rank these days is very simple, but you have to get out of the "old days" mentality.

1. Don't write the articles around keywords, but rather a topic. Pages that I rank have 0% keyword density - that's right - 0%. But they are all themed properly. Use tools for that (I won't promote them here, but you know which ones i am talking about). I'm not here to preach some white-hat nonsense, I'm telling you this from my own experience. Google is moving away from the keyword driven queries, and moving into the field of interpreting queries. Hummingbird was the first step.

2. Try to sculpt the Pagerank on your site by creating thematically tight categories. You could call it siloing, but I wouldn't go that far anymore, since I've seen some of the sites tank if I tightened the silo too much. I'm not talking about the browser Pagerank, but rather their internal Pagerank, which is nothing but a simple linking stream of your site.

3. Hire a good UI designer, or try to do it alone. I'm a student of multimedia engineering, and my college collaborates with Google on some projects here in my country. You wouldn't believe what you can find out if you can read between the lines. People are saying that the design doesn't matter and that links are all that matters, but I can assure you, this couldn't be farther from the truth. Sure, they can't read the design directly, but they can read it by data that the user gives them (even if he/she isn't logged in). Bounce rate is just a drop in the sea here, and I'm not talking about the flawed bounce rate metric in the GA. I'm talking about the real bounce rate.

All in all, if your site doesn't rank well, and you don't know why, read the points above. Also, keep in mind that some of the penalties and triggers are much more complex than what people are discussing on blackhat forums. For example, I know for a fact (don't ask me how I know this) that Panda and Penguin are working in tandem; this means that if your site is targeted by Panda (and most of the sites owned by blackhatters are, they just don't realize it, and think that it's Penguin, simply because their "GA bounce rate" says it's "good") every bad link that you point to your page increases the chance of triggering the Penguin exponentially. Although they are not directly connected, you can see where this is going.

Anyways, this has been my rant.
 
Cool study man! One question: did you do anything to help the links get indexed?

We would normally index links for our own sites but for the purposes of this test, we didn't even ping the links. We just let Google find whatever it was going to find naturally over time.

Why not run the successful test in combo towards the "safe" video and see what you get?

Why not try and rank an article on blogger or some other site the same way?

Please make useful test and don't lead the newbies astray... Now they are going to go away from some of the links that "didn't work" when if you ran do follow wiki's and got them indexed you could have ranked your video - done it before for my own test. And yes - there are do follow wiki's....

1. Because if we discussed every factor and authority of this video, it would be a 2 hour video. You're asking for an anal level of detail instead of just being grateful for the value that's already been shared.

2. We made a case study to rank something that made actual money without intending to and then shared that with the community. I fail to see how that steers newbies in any wrong direction. You'd think this was a WSO that promised to make $100k in your underwear from pushing a button twice per day reading your reply. If anything, a newbie could read between the lines and figure out how to rank a video for a decent keyword and have it generate income. That's useful to a lot of people here.

That's interesting that you said that. I've always thought that because SEO is so transient, most people don't have time to learn why their methods work, so they just use those methods as a short-term strategy without completely understanding them.

I understand his thinking but I would highly disagree.

It sounds good in theory but how would he possibly know why Google chooses to value some links over another? Does he secretly work for Google? Is he a spy in their algorithm? What secret method do you suppose he has to figure out WHY Google ranks some things and not others? The truth is that the general public doesn't have access to that information.

In reality, educated testing is all we have to make the system work in our favor and make money.

I think people tend to forget Google is a mathematical algorithm, not a person who chaperons over your website.

Finally something useful shared. Yeah,short term seo is most less risky in terms of investment/roi wise.

It would ve been better if you had done PR and articles separately to see actually which one brings juice. :cool:

Good point. Yeah what was my fault definitely. I thought originally that we made 10 videos and had 10 different things to test. At the end, I left one as the test video untouched and then realized we only had 9, so I had to double up on one of them. I can tell you that it was 99% articles and 1% PR so it's more than likely the articles doing the bulk of the ranking.


Here's a hint for you guys: +1s, tweets and likes don't work for SEO, but social shares (when somebody actually links to your post from Google+ or Facebook, or shares it on his/her own wall) actually work wonders.

Also, I see so many people tackling backlinks and Penguin, when in fact their rankings are plummeting because of Panda...yes, that's right, PANDA, not PENGUIN. I can only tell you that Panda is mostly driven by user search metrics (Google owns Chrome and has invested a lot of money into Firefox, A.K.A - they have the backend data on Firefox users as well) - hire a UI/UX designer make the user experience the best possible, so that the user metrics increase in your favor. There are some tools out there which you can use to manipulate these metrics.

Forget about backlinks, you only need around 10 of them which are relevant (hint, a private blog network consisting of expired domains with high DA, PA and Trust flow without a long record of being expired)- that's enough for the backlinks. Then focus on social, but not those crappy likes and +1s, but on real social link sharing on people's own walls. Finally, focus on good user experience to increase your user metrics stats, or manipulate it.

The way to rank these days is very simple, but you have to get out of the "old days" mentality.

1. Don't write the articles around keywords, but rather a topic. Pages that I rank have 0% keyword density - that's right - 0%. But they are all themed properly. Use tools for that (I won't promote them here, but you know which ones i am talking about). I'm not here to preach some white-hat nonsense, I'm telling you this from my own experience. Google is moving away from the keyword driven queries, and moving into the field of interpreting queries. Hummingbird was the first step.

2. Try to sculpt the Pagerank on your site by creating thematically tight categories. You could call it siloing, but I wouldn't go that far anymore, since I've seen some of the sites tank if I tightened the silo too much. I'm not talking about the browser Pagerank, but rather their internal Pagerank, which is nothing but a simple linking stream of your site.

3. Hire a good UI designer, or try to do it alone. I'm a student of multimedia engineering, and my college collaborates with Google on some projects here in my country. You wouldn't believe what you can find out if you can read between the lines. People are saying that the design doesn't matter and that links are all that matters, but I can assure you, this couldn't be farther from the truth. Sure, they can't read the design directly, but they can read it by data that the user gives them (even if he/she isn't logged in). Bounce rate is just a drop in the sea here, and I'm not talking about the flawed bounce rate metric in the GA. I'm talking about the real bounce rate.

All in all, if your site doesn't rank well, and you don't know why, read the points above. Also, keep in mind that some of the penalties and triggers are much more complex than what people are discussing on blackhat forums. For example, I know for a fact (don't ask me how I know this) that Panda and Penguin are working in tandem; this means that if your site is targeted by Panda (and most of the sites owned by blackhatters are, they just don't realize it, and think that it's Penguin, simply because their "GA bounce rate" says it's "good") every bad link that you point to your page increases the chance of triggering the Penguin exponentially. Although they are not directly connected, you can see where this is going.

Anyways, this has been my rant.

Good rant. I would also add that you should document all your links you build so you can later determine what links you need to remove if necessary. None of the backlink tools will find all of your links so you need to know how to fix them sometimes. Maybe I'll add another thread on how to tell if your links are penalized for people to remove their bad ones later.
 
The video was great. I'm going to do my own test out of pure curiosity, now. Thanks!
 
Back
Top