The Importance of Benchmarking
You're about to launch your brand-spanking new website, and maybe you've even spent time (or money) getting your SEO ducks in a row. How will you know whether or not your efforts have paid off in terms of more people finding you? And sticking around longer? And perhaps even donating more?
The way you know is called "benchmarking". Benchmarking is the process of looking at current metrics, so that you have a way of knowing truly whether or not what you've done has made a difference.
I've been doing this kind of stuff for a while now, and part of it is fun, and some is grunt work. But it's worth it. Some of the benchmarking (like measuring how long people spend on your site, and how many pages they visit) is easy with good web metrics software like Google Analytics. But the most important thing is: don't wait until you've launched your new site to start measuring - start before, if you can. Install GA on your old site a month or two before launch, so you'll have a baseline of measurement to go from.
The kind of benchmarking that is grunt work is finding out where your site comes up on search engines based on the keywords you want people to find you with. You should do searches with at least 10 important keywords or phrases, with at least two of the major search engines (one of which, of course, starts with "G".) You certainly can outsource this kind of work - it's pretty straightforward. But it is also important to do this task with a bit of knowledge about the organization - it's good to know where you show up in relation to organizations like yours. I've done a couple of projects where with some keywords, organizations came up higher in searches that had less domain expertise than the one I was working with. It's good to know these things going into a new website launch and SEO process.
And then, the idea is to do the same benchmarking a month or so after site launch, and see how your site rebiuld and SEO efforts fared, and see what tweaks need to be happen. And keep benchmarking and tweaking as time goes on.
The way you know is called "benchmarking". Benchmarking is the process of looking at current metrics, so that you have a way of knowing truly whether or not what you've done has made a difference.
I've been doing this kind of stuff for a while now, and part of it is fun, and some is grunt work. But it's worth it. Some of the benchmarking (like measuring how long people spend on your site, and how many pages they visit) is easy with good web metrics software like Google Analytics. But the most important thing is: don't wait until you've launched your new site to start measuring - start before, if you can. Install GA on your old site a month or two before launch, so you'll have a baseline of measurement to go from.
The kind of benchmarking that is grunt work is finding out where your site comes up on search engines based on the keywords you want people to find you with. You should do searches with at least 10 important keywords or phrases, with at least two of the major search engines (one of which, of course, starts with "G".) You certainly can outsource this kind of work - it's pretty straightforward. But it is also important to do this task with a bit of knowledge about the organization - it's good to know where you show up in relation to organizations like yours. I've done a couple of projects where with some keywords, organizations came up higher in searches that had less domain expertise than the one I was working with. It's good to know these things going into a new website launch and SEO process.
And then, the idea is to do the same benchmarking a month or so after site launch, and see how your site rebiuld and SEO efforts fared, and see what tweaks need to be happen. And keep benchmarking and tweaking as time goes on.
Labels: seo
Subscribe through Feedburner
1 Comments:
I think you have to be careful about putting too much stock in search results positions.
At least for Google, search results are personalized based on account history with their "advanced personalization technology" [1]. So if you often search for your organization, results related to your organization may get a higher search ranking just for you.
Results are also different for people searching from different locations. Google notes that "The same query typed in multiple countries may deserve completely different results" [1].
Also, Google does many experiments on users by altering results for some users and then measuring how the altered group reacts (in terms of clicks, time spent, etc). This is done to gather data to use to improve search relevance [2].
So I'd make sure to note the circumstances of the original search tests and try to replicate those circumstances for the later trials. Even that could be pretty inaccurate though, if you're chosen to be in a search results experiment for example. I'd pay more attention to how many visitors to your site come through particular keywords.
---
[1] http://googleblog.blogspot.com/2008/07/technologies-behind-google-ranking.html
[2] http://news.cnet.com/8301-10784_3-9954972-7.html
Post a Comment
<< Home