Jump to content

  • Log in with Facebook Log in with Twitter Log In with Google      Sign In   
  • Create Account

Subscribe to HRA Now!

 



Are you a Google Analytics enthusiast?

Share and download Custom Google Analytics Reports, dashboards and advanced segments--for FREE! 

 



 

 www.CustomReportSharing.com 

From the folks who brought you High Rankings!



Photo
- - - - -

Keyword Research From Your Logs


  • Please log in to reply
16 replies to this topic

#1 jsp1983

jsp1983

    HR 3

  • Active Members
  • PipPipPip
  • 98 posts
  • Location:Blackpool, UK

Posted 29 April 2009 - 11:01 AM

Something I've been meaning to ask for a while and it's a bit of a basic question, but one that's never really been explained to me.

Quite a few articles I've read have recommended that you trawl through your visitor logs and analytics stats to find the keywords and phrases that visitors are using to get to your site. Easy enough to do, but what I've never seen explained is 'why?'.

I have a couple ideas...

- To find phrases that could be made more prominent or focused on particular pages to increase traffic and/ or increase conversion quality.
- To find phrases that you aren't optimising for, but which your page is ranking for, meaning if you use the actual phrase, you might rank higher in the SERPs.

Are they even close?

#2 Randy

Randy

    Convert Me!

  • Moderator
  • 17,540 posts

Posted 29 April 2009 - 11:05 AM

They're not only close, they're right. appl.gif

There are some additional ones. Starting with when you see traffic showing up for phrases for which you've not actually optimized those can then lead you to even more phrases for which people are searching that fit your site well. As a for instance, let's say you see someone or several someones reaching your site via one or more 3 or 4 word phrases. If you strip that down to a two word phrase (usually more competitive) that you're not currently optimizing for, you may want to start optimizing for it to pick up an additional traffic channel.

Make sense?

#3 jsp1983

jsp1983

    HR 3

  • Active Members
  • PipPipPip
  • 98 posts
  • Location:Blackpool, UK

Posted 29 April 2009 - 12:35 PM

QUOTE(Randy @ Apr 29 2009, 05:05 PM) View Post
They're not only close, they're right. appl.gif

There are some additional ones. Starting with when you see traffic showing up for phrases for which you've not actually optimized those can then lead you to even more phrases for which people are searching that fit your site well. As a for instance, let's say you see someone or several someones reaching your site via one or more 3 or 4 word phrases. If you strip that down to a two word phrase (usually more competitive) that you're not currently optimizing for, you may want to start optimizing for it to pick up an additional traffic channel.

Make sense?


Umm, yeh, I think so. Could you give an example, just so I know I've understood you correctly?

Now, I was thinking about how I might go about trying to see which phrases I've included on a site and which ones I haven't (say, for example, I've been a bad SEO and not kept track of phrases I've used).

I know I could use the 'site:' operator on Google manually enclose each phrase in double quotes to see if the phrase is on indexed pages, but this seems a bit slow and tedious. Is there a method or a tool available where I can drop in a list of phrases and search for them on indexed pages all in one go?

Don't get me wrong - if I have to do it the hard work way, I will, but like you sig says, Randy... Why take the risk? tongue.gif

#4 Jill

Jill

    Recovering SEO

  • Admin
  • 32,979 posts

Posted 29 April 2009 - 01:23 PM

You might want to have a read of my recent SEL article:

Using Google Analytics to Measure Your SEO Success

#5 Randy

Randy

    Convert Me!

  • Moderator
  • 17,540 posts

Posted 30 April 2009 - 03:10 PM

Sure, I can give you an example jsp. A real one in fact, though it's on my personal blog and not on an e-commerce site. Not that it couldn't be an e-commerce situation. Given the amount of traffic I get around this one little phrase I could probably start a business installing a free package on people's servers. lol.gif

I did a post on my blog some 2 1/2 years ago explaining how to install a server software package named ImageMagick on a Redhat Linux server. Why I made the post originally is that IM's default install doesn't support some image filetypes. So my little tutorial was basically showing folks how to expand its capabilities to support more image filetypes. Pretty simple stuff, right?

Aside: I'm not linking to the post here. It's rather old though still works with updated libraries, plus it's easy enough to find if someone stumbles across this thread when it starts ranking for these phrases and needs the full tutorial.

The original post had a title and headline (it's a movable type blog, so they're the same) of something like Installing ImageMagick on a Redhat Server Sure that explained the post, but not really. It was too generic.

When I bothered to look in my stats I saw I was getting small traffic numbers on all sorts of related, but more focused phrases. Phrases that mentioned an image format someone needed IM to work with that it didn't out of the box. Phrases like: imagemagick jpeg, imagemagick jpeg support, tiff support in imagemagick, jpeg imagemagick, png support imagemagick, etc. The common denominators being the imagemagick name, which is basically a brand name, one of the popular image filetypes and often the word support.

When I looked up several of these more focused phrases on Google, yes I sort of ranked for all of those 20 or 30 variations. But I didn't rank great. Mostly on page 2 or 3. All of the words were on the page, just not in the same order people were always searching for them. So people were having to dig pretty deep to find my little post. So I was getting just a trickle of traffic (1's and 2's and 5's) on each of these phrases. Several of those folks that did find my post left me comments or even sent me email thanking me for the tutorial.

So me being me, I decided to make a little tweak to the page so it would show up better for these extra phrases, based upon nothing more than on-page factors. Basically I did the same thing I've been doing with e-comm sites and their long tail phrases for years. All I did was change the title of the post, which in turn updated the <title> and headline on the page. A simple change really. The post title became: Installing ImageMagick with JPEG, TIFF and PNG Support

As I said above, this isn't an e-commerce site. And the ancillary phrases are not terribly competitive. Frankly, it doesn't hurt me one bit if the blog post doesn't rank, so it's not something where I'd go out of my way to create new pages to target all of these possible phrases or start a mini-link building campaign or anything like that. But I've found often tiny changes like this work on e-comm sites also. It's where I always start, since it's the easiest to accomplish.

Turns out I didn't have to do anything extra. My little blog post has ranked pretty much right behind IM's own site (which is THE Authority site but doesn't explain how to do this sort of thing) for all of those 20-30 phrases. It's been there at the top and stuck there for over 2 years, even though there are others out there who do sell a service to install IM with these extras.

The moral being I'd have never known people were searching for those phrases without picking upon the little trickle of traffic in my stats. The phrases don't make it to anybody's radar screen. Even Google's keyword tool shows each of them being searched on less than 100 times per month. They're not worth shooting for in those low volumes, right? Wrong. Together these little easy-to-rank phrases add up to several thousand visits per month. Every month. With basically zero effort on my part. And it's been the same for over two years now. Not to mention the searchers are looking for exactly what I'm "selling".

That's the lesson. Too many people with e-commerce sites pass up really easy to capture traffic like this. They get so focused on one or two Ego phrases (that usually don't convert worth a crap I might add) that they totally miss the dozens and dozens of closely related, more relevant and better converting phrases they could capture with minimal effort. Nobody else seems to find them because they don't cause a blip on during initial keyword research, and apparently few actually pay attention to their stats.

#6 Jill

Jill

    Recovering SEO

  • Admin
  • 32,979 posts

Posted 01 May 2009 - 05:18 PM

yes.gif

And that's exactly what that SEL article I mentioned above as well as the article from this week's newsletter was getting at as well!

#7 jsp1983

jsp1983

    HR 3

  • Active Members
  • PipPipPip
  • 98 posts
  • Location:Blackpool, UK

Posted 08 May 2009 - 09:02 AM

Thanks for the example, Randy cheers.gif

I've spent all week agonising over keyword management... I'm sure I'm not meant to be so scrupulous about it! I think I've perhaps gone to the opposite extreme of what you mention about people focusing on one or two keywords. I'm finding myself trying to find ways of prioritising relevant keyphrases (as per your PM) I've found with extremely low monthly search volumes (in the tens), alongside the ones with thousands of monthly searches- that's before I've even started to deal with phrases found from my logs.

One thing I've picked up on in reading people's guides to keyword management is the recommendation for systematically using modifier words like 'buy', 'cheap', 'guide' etc. I've found that these sorts of words haven't even shown up in my keyword research (or at least, their search volumes in the Google Keyword Tool says 'not enough data'). Is this a method that's more suited towards broader-themed sites than niches?

The SEL article pretty much reinforces the point, although I've yet to set up GA properly to track conversions etc.

One thing I have noticed about GA is that it doesn't appear to track as many keyphrases that have referred traffic to my site than Webalizer does.

#8 Jill

Jill

    Recovering SEO

  • Admin
  • 32,979 posts

Posted 08 May 2009 - 10:08 AM

QUOTE(jsp1983 @ May 8 2009, 10:02 AM) View Post
One thing I have noticed about GA is that it doesn't appear to track as many keyphrases that have referred traffic to my site than Webalizer does.


Then you must not have something set up right because it should track all keyword phrases that people have found your site from.

#9 Katy

Katy

    HR 3

  • Active Members
  • PipPipPip
  • 68 posts
  • Location:Oxford, England

Posted 08 May 2009 - 10:36 AM

QUOTE(jsp1983 @ May 8 2009, 02:02 PM) View Post
One thing I have noticed about GA is that it doesn't appear to track as many keyphrases that have referred traffic to my site than Webalizer does.


I've noticed there's quite a big discrepancy in what my logfiles say compared to what GA says in terms of keywords and referring sites with my log files capturing a lot more data than GA does. I put this down to three reasons (although feel free to shout at me if I'm talking a load of rubbish):

1) some people will have javascript disabled so GA can't track them
2) GA experiences an intermittent "blip" so doesn't track the data
3) As the Log files track direct hits to the server - page/image retrevials etc it's a bit more accurate about what it can track

just my 2 penneth worth!

Katy

#10 Jill

Jill

    Recovering SEO

  • Admin
  • 32,979 posts

Posted 08 May 2009 - 11:31 AM

You are definitely correct, Katy.

I wouldn't think the differences would be extreme, but I've never really compared numbers.

#11 jsp1983

jsp1983

    HR 3

  • Active Members
  • PipPipPip
  • 98 posts
  • Location:Blackpool, UK

Posted 08 May 2009 - 01:04 PM

The differences aren't extreme, but I have noticed the discrepancies. Not enough to cause a worry, but just an observation I've made.

GA is most definitely set up correctly - I use the Headspace plugin for WP and from what I've seen, there are no complaints about it incorrectly tracking with GA.

I'm inclined to go with Katy's first suggestion as the reason for the slight difference. I'm amazed that people still disable JavaScript these days, but that's just my own PoV.

#12 Randy

Randy

    Convert Me!

  • Moderator
  • 17,540 posts

Posted 08 May 2009 - 01:31 PM

In some cases it's not even Javascript being disabled.

I know of at least one place that recommends disallowing google-analytics.com via the hosts file on the theory that The Man is tracking too many people and gaining too much data. The paranoia I've read doesn't even suggest Google is doing anything nefarious with the data. It's more worry that some government is or has forced them to hand over such personal surfing history. What they never seem to take into account is this type of action would break the trust Google has built up with users over the years. So even if the subject has been broached, and I'm sure it has, it's something they're going to fight everything they have and make sure any loss on their part becomes very public so that they don't get the brunt of the blame.

Blocking a domain like that is an easy thing for folks to do and frankly doesn't cost them nearly as much enjoyment if they simply restrict that kind of data pass through when compared against disabling JS altogether. So I can see why someone with a bit of paranoia might consider it.

I'm careful, but not quite that crazy. My feeling is if you're that paranoid you probably shouldn't be on the 'Net anyway, since someone else is always going to have ultimate control over this type of data. And you can never truly know how they might use it, who they might sell it to or give it to.

#13 jsp1983

jsp1983

    HR 3

  • Active Members
  • PipPipPip
  • 98 posts
  • Location:Blackpool, UK

Posted 08 May 2009 - 02:42 PM

Well, I've covered my laptop in tinfoil, so I'm protected from the Google... Aren't I? ninja.gif

#14 Randy

Randy

    Convert Me!

  • Moderator
  • 17,540 posts

Posted 08 May 2009 - 07:31 PM

Nope.

Google patented and developed a tin foil penetration algorithm back in 2001.

jester.gif

#15 jsp1983

jsp1983

    HR 3

  • Active Members
  • PipPipPip
  • 98 posts
  • Location:Blackpool, UK

Posted 20 May 2009 - 07:09 PM

Okay, so, I've been going through my logs to see which phrases have been referring traffic. Where it's just a single occurence, I've disregarded it - all others I want to look into a bit more.

Now, I'd like to know if there's a tool that will help me to find instances of phrases on my site. I know that I could use the site:[URL] "[keyword phrase]" operator on the search engines, but not all of my pages are indexed yet (I know - why bother if the page isn't yet indexed by an engine, but some pages are indexed by Google and not Yahoo! etc). Is there a tool available that will search through all of my site and will look for a particular phrase or set of phrases?

And just coming back to what I asked in my previous post...

QUOTE
One thing I've picked up on in reading people's guides to keyword management is the recommendation for systematically using modifier words like 'buy', 'cheap', 'guide' etc. I've found that these sorts of words haven't even shown up in my keyword research (or at least, their search volumes in the Google Keyword Tool says 'not enough data'). Is this a method that's more suited towards broader-themed sites than niches?


Could anybody share their thoughts on this?




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users

SPAM FREE FORUM!
 
If you are just registering to spam,
don't bother. You will be wasting your
time as your spam will never see the
light of day!