I’ve been using Google Website Optimizer (GWO) in one form or another for over a year now, and I am quite impressed. I wish everyone would take a look at this technology.
The short version of this post is: if you have a website that gets some traffic, and actions that you want a user to take, such as buying something, you should use GWO.
Let’s talk about GWO a bit more for everyone who didn’t leave after the executive summary.
First, what is GWO?
It is a free add in to your website that lets you test the effect of content changes on desired user actions. For example, you can test the impact of different headlines (‘Sign up for more information’ or ‘Free Newsletter!’, etc) on a newsletter signup page to see how it affects how many users sign up.
I could try to write about it more, but instead, here’s a video that explains how to set up a test; you can watch the first 45 seconds to get a good overview of what GWO helps you test, but the whole video is worth viewing.
Now, how should GWO be used?
This tool should be used to optimize a specific action that you want the user to take, and that action should be something on the web. Examples of actions that GWO can help optimize include:
- sign up for a newsletter
- visit a specific page
- order something
- give you their contact info
- watch a video
- download a file
Next, why would you avoid GWO? There are a number of scenarios where using this tool just doesn’t make sense.
No actions that you want the user to take on the web
This is the biggest set of sites for which GWO does not make sense, and my blog falls into it. My end goal is really to talk to someone, on the phone or over coffee, to ascertain whether or not I can help them with my services. That’s a hard conversion to track. If you really want to get people talking on the phone, that is a conversion that GWO doesn’t help track either (though it’s possible you could do something with VOIP or different phone numbers and cobble some kind of conversion tracking together).
That’s not to say that sites in this category can’t use GWO. For instance, I wrote an article long ago about a java technology called JAAS, and it’s still pretty popular. I’ve been doing some testing on that page to try to drive traffic to this blog, where a big chunk of my current writing is happening. I was able to increase conversion about 2.5 times (granted, from a small base–0.7% to 1.73%). I could definitely try to do some optimizations around ad placement or contact calls to action. The main reason I haven’t is that I still believe my best chance at conversion is referral or word of mouth, not better text on a website.
Google trust issues
For providing this valuable service, Google could get a look at a lot of interesting data including how many websites optimize, what site admins are tracking, and what URLs those admins consider important. I have no idea what Google actually does with the data–the company may not track the data at all. I did a bit of looking on the GWO forum, but didn’t find anything useful. However, if you don’t want Google having access to this data, then you shouldn’t use GWO. I don’t use Google Analytics for this very reason.
Speed is essential
Not enough traffic to your site
If you don’t have much traffic, your effort will be better spent driving more people to your website. Whether you use search engine optimization, TV advertising, pay per click campaigns, or my favorite, regularly posted organic content, traffic is a prerequisite to using GWO effectively. How much traffic? Well, I was able to get conclusive results in my JAAS experiment with 3466 views and 48 conversions in just about 3 months. That’s about 40 views a day. I was not testing a ton of variations in my content (only 3).
Not enough time
I don’t believe this is a valid excuse–spend 7 minutes and watch the video; I believe you’ll see that this is worth a bit of investigating if the above scenarios don’t apply to you.
I can’t think of any other reasons not to use GWO. Feel free to leave a comment if you feel I missed something.
What are your next steps?
Well, if I’ve convinced you that GWO is worth looking at, you might want to do a bit of further research. Here’s the website optimizer homepage. Here’s the official GWO blog. Here’s a great article full of tips for GWO. There’s a book called Always Be Testing that seems pretty well received that you might want to check out. I have not read it.
Disclaimer: I’d love to help you out with this (if interested, contact me), but firmly believe that almost anyone can take 30-45 minutes and implement this on their website. I think GWO is great, cool, useful technology and want to see it used more to make the web a better place.
In regards to GWO not working on sites like WordPress, there actually is a plugin which allows you to use GWO on a WordPress site: http://websiteoptimizer.contentrobot.com/ With the number of plugins available for WP, I actually find it to be a very flexible platform to build a site with. My site has a huge amount of content and interactivity, and it has scaled very well.
Thanks for the article, Dan!
That’s a great plugin for WordPress–thanks for posting it.
My point above is that WordPress.com (which provides excellent free blog hosting) doesn’t allow you to use GWO (via manual template editing or that plugin).