Grow Your List Through Testing with Google Website Optimizer
Monday, August 31st, 2009Howdy folks, and welcome back. Today we are going to attempt to increase the amount of people signing up for your email list through a little bit of testing using Google’s amazingly intuitive (and free) Website Optimizer. To execute this, you need a Google account, and you must be able to both make visual changes to your site and add code to your HTML. If you have an internal web team, work with an interactive agency (like us), or a freelancer, the changes shouldn’t be too difficult to for them to execute.
A/B Experiment Checklist
There are many elements of an opt-in form that can dramatically affect the conversion rate. For sake of this post, we will choose placement on the home page of the opt-in box. The theory is that if you give more promance to your email opt-in, you will get more sign ups. Following thse steps will allow you to find out if this is true for your audience.
- Choose the page you would like to test
- Create alternate versions of your test page
- Identify your conversion page
To test our thoery, create an alternate version of your exisiting home page, with the opt-in box in a different location. For the page variation, use a name like site.com/index2.html. The page will not be accessible to users unless it is served up by the page loader script.
The conversion page is very important in the process, as this is the validation of a successful visitor. You want to make sure that your thank you page is strictly for your email marketing and not also used by other forms. Also, if you have a double opt-in process, make sure that you idenitfy the first confirmation page as your conversion page, not the page a vistitor goes to from the confirmation link in an email. Getting them to opt-in is the point, not email confirmation.
Installing and validating JavaScript tags
This is where things stop getting polite and start getting real technical. If you are unsure about adding code to your website, forward this link to your web team and they should be able to do it in minutes.
There are 2 scripts, - a control script and a tracking script. The control script should appear immediately after the opening <head> tag of the original page.This is the script that communicates with Google’s servers to retrieve alternative page information, and ensures that individual users are tracked properly, by showing them the same variation each time, and by not double-counting their visits should they come back to the page at a later time.
The tracking script is pasted on all 3 pages directly before each page’s closing </body> tag. This script sends pageview information to Google, so that visits will be recorded in your reports. The nice thing about this tool is that it validate the scripts before activating the test and also gives you links to send instructions directly to your web team.
May the Most Conversions Win
One thing I love about testing is that no one is right or wrong until the results are in. You can best practice and benchmark your site to death. But you will never innovate until you test your theories. If you have an idea and a web designer tells you it’s no good – Test it! If you think a different color button will get clicked on more – Test it! It’s really fun to watch the results come rolling in. The Google Website Optimizer will declare a winner, but it needs at least 100 conversions, so if you don’t get a ton of traffic, you may need to leave it up for a few weeks.
Testing placement is just one area to test. The goal here is for conversions. Anything that relates to getting information or purchases from users is on the table. Do you have any elements you have tested that provided big results? Let’s hear about them in a comment below!
Our first guest is George Bilbrey (
