Are you a Google Analytics enthusiast?
More SEO Content
Google Analytics Data Sharing
Posted 07 April 2008 - 05:51 PM
Take my personal site for instance. Much of its search driven traffic comes from all things ImageMagick because I have a couple of blog posts up about how to install and configure this particular server software. But I also get traffic on keywords like Common Link(s) because that's what I named one little tool, .htaccess stuff, css stuff (I built a css rollover tool years ago ), Yahoo, MSN and Google API phrases. Those and others I get traffic on could all fit in a server/programming/IT niche I guess.
But if they're using solely keywords how are they going to justify putting my site in a server/programming niche when it also pulls traffic for phrases like chatterbox (it's the name I gave the blog long ago), conversion analysis stuff and <gasp> the dress designer Alfred Angelo? Don't ask me how I get traffic on that last one. I've never bothered to look, but I assume I mentioned Alfred's name and dress styels in some code I posted on the site some time or the other, because I do get a little traffic for some specific dress styles he sells.
Let's just say there has to be more to it, or perhaps a breaking point somewhere. Because I don't see how they could figure out what niche my blog is supposed to be in. Especially since I don't know what niche it's supposed to be in!
It would certainly be nice if they'd give a few more details about how it all works. Nothing trade secret-y mind you, just some additional general info.
Posted 08 April 2008 - 04:59 AM
I've got the same large range of different goals, some where a pageview is a goal and others where a goal is an actual sale that happens once every blue moon. Could be that if you have a large enough sample size it all calculates ok.
Wouldn't it make more sense to determine the vertical by the keywords on the pages itself? Don't know whether they would use a separate crawler or the standard google one.
Posted 08 April 2008 - 06:27 AM
The math would be mind numbing if you consider the concept of providing benchmarking totals for whole sites where the verticals/niches are determined on a page-by-page basis, however since Google does everything else on a page level it should be a piece of cake for them. Or they could simply look at all of the keywords from which a site attracts traffic and decide that since 60% fit in one niche, 25% fit in another and 15% in another, they'll benchmark it in the largest. The could even look at gross traffic or sheer number of keyword phrases that fit in either.
This is the kind of general informaiton I'd like to see them release so that everybody has a better understanding of what we're seeing and not seeing in the numbers.
Posted 08 April 2008 - 06:44 AM
Posted 16 April 2008 - 10:21 AM
Very valid point. I was actually thinking of accepting the terms for a site and then having goals set up with a high and low conversion values with the goal being the actual home page itself. About an 80% conversion rate should look great in some respects surely?
Hmmm, Then again, maybe G will then up my ad prices 5 fold as it knows Im in making a profit situation for certain phrases. Thats if the goals are set up correctly and showing this info. (personally I dont add genuine goal value to my analytics. I add values such as 1c and 3c that I can relate to the actual real situation. I really dont want google, or anyone else for that matter knowing how much money I am making)
Way way too generic are these types of tools for me. I want to measure 4 or 5 totally different niches, to do that Relevancy has to be spot on. Google cant even get the relevancy themed sites particularly great in the serps where im still considered relevant/similar to monkey training for a restaurant site. Why/How is this going to be any better than that? or better than alexa or compete or any of the other tools that we may combine currently for accurate (all relative) feedback.
To be honest. I dont really get into checking all my competitors to a major degree. I worry about the percentages, bounce rates, conversions on my own sites and aim to improve all of those aspects, keep the graph on an upward curve so to speak. Having google tell me that generally everyone in my generic niche is getting 5 times as many page views or 1000 times more traffic really doesnt mean anything in that generic form.
The only real benefit I see would be to see an individual site with great conversions at a profit making situation. You look at the site, the page is unusual layed out or written so clever that I may want to incorporate element into my own 'sales pitch'.
But as we are saying, that is not going to happen now, or ever in my opinion.
This actually reminds me of the Marketleap tools i used to use a few moons ago, it used to incorporate some generic benchmarking guides.
Posted 16 April 2008 - 10:56 AM
Amazing. That's the first thought that crossed my mind when I looked at the data. I guess that great minds thinking alike axiom is true, hunh?
Seriously now, as things stand today the only advantage to accepting benchmarking that I springs to mind is in the very rare case where you were trying to sell a domain. Especially an e-commerce domain. I can't believe that the massive skew I'm seeing, where my two test sites are waaaaaaay over the norm is an isolated circumstance.
With the numbers it shows me and how I'm apparently seriously outperforming my "competition" in every single area, providing the data might help overcome any reluctance on the the part of the buyer. Problem is even if I were going to sell off some sites I'd never use that GoAn Benchmarking data as part of the sell/close because I know without a doubt that the numbers contain some serious skew.
I guess I simply need to lose my morals, sign up more of my sites and start selling these things off for an inflated sum. I don't doubt someone will be doing this. Next week probably.
Posted 16 April 2008 - 06:28 PM
Posted 16 April 2008 - 09:37 PM
Why do I get this mental image of a Darth Vader character, cept with a blue G on his cape, uttering something about being my daddy.
Posted 17 April 2008 - 05:21 AM
That would be really weird, if only for the fact that you are older than G. :insert time traveling emoticon:
Posted 20 April 2008 - 08:54 AM
Dunno if this is new or I simply hadn't noticed it before, but now the Benchmarking thingee gives you the option to see and set a general vertical category to check your site data against.
From what the Help Docs now say it looks like their initial categorization is more along the Size of the site --with a very general Small, Medium and Large being the option they choose-- as opposed to being a business-type category. I'm reading what they say in the Technical Details section of the above page to come to this conclusion, though I am reading between the lines a bit. Also, it appears this Size categoration isn't really about the number of pages a site has or the number of pages indexed, but instead is apparently triggered by the amount of traffic the site gets.
Once you're logged in you can choose a vertical business type to check against, however you apparently can never compare outside of sites of a similar size. You can only compare your site against one type of site vertical at a time, but it's immediate so easy to do. The category list I get on both of my test sites is identical, so I assume everyone will have the exact same categories to choose from. There are roughly 30 major categories, with most of them having at least a couple of more specific sub-cats.
So to round out what I'm seeing this morning...
- Initially your benchmarking is compared to site in any subject category that are roughtly the same size as your site.
- The size is a very general Small, Medium and Large, with no solid definition of what this means. The helps docs hint that "size" speaks to the number of visits a site receives, not the number of pages that are indexed. Odd that considering they use the word size, but oh well.
- You cannot manually change what Size cat your site falls into.
- Once you start getting data you can compare against sites of a similar size that have been determined to be in one of several dozen vertical categories.
- Each category is required to have at least 100 accounts who have benchmarking enabled. Note it's Accounts, not Sites. So for someone like me who has 2 sites in a single account, if both sites were in the same category (mine aren't, but I'd bet many are) it would only count as 1, not 2 or more.
- When you check against a vertical category the output changes immediately. So you can check against several cats, even though you can only check against one at a time.
- You can also go back to the default of All sites of a similar size.
Posted 21 April 2008 - 03:07 AM
It was there the first time I had data in the benchmarking.
0 user(s) are reading this topic
0 members, 0 guests, 0 anonymous users