Google Website Optimizer v 1.0 The Techie Guide - Manual (Page 5)

5
The Techie Guide to Google Website Optimizer
How Experiments Work
Depending on how you set up your experiment, each visitor could see one of a dozen or more combinations
of the headline, image, and call-to-action button on that page.
In other words, one would see the puppy image with headline A and call-to-action button B. Another would
see the kitten image with headline C and call-to-action button A. And so on.
Let’s take a closer look at an MVT experiment in action, continuing the three-section experiment we
described above. Here’s the sequence of events:
1. A visitor lands on your test page: http://www.mysite.com/landing_page.html.
2. The GWO Control Script (which is JavaScript code) executes, which generates a request to the GWO
servers to request an additional JavaScript
known as siteopt.js.
3. If this is the first time the user has seen
this experiment, siteopt.js sets cookies to
note what variation of the page the user has
seen, and it sets up variables and defines
the utmx_section function so that later in
the page, as the document is parsing and
executing, the code there will make use of
the variables to replace the section content.
If this is not first time the user as seen the
experiment, then siteopt.js will read the previously created cookies. It will then define the utmx_
section function such that later in the page the user will see the same variation they saw when they
last saw the experiment. Additionally, the cookie expiration date will be reset to two years.
4. A third piece of code, which we call the Section Script, determines which section variations will
appear to users. As the browser parses the page, it will encounter Section Scripts, which can modify
the original content. It calls utmx_section function that was defined by siteopt.js earlier in the page.
5. The Section Script works with siteopt.js to replace the original version of the content with an
alternate version that includes alternate headline B, a kitten image, and call-to-action button C.
Some users, chosen at random, will see the original version of the page as a control. In this case, the
Section Script will leave the original content in the page unchanged.
6. The Tracking Script on the test page uses the cookies set by siteopt.js to note which combination
of content was displayed to the user. The Tracking Scripts submits this information to GWO for the
purpose of reporting.
7. If the user (we hope) then visits your
conversion page, such as http://www.mysite.
com/thank_you.html. The GWO Conversion
Script, which is nearly identical to the
Tracking Script, records the visit to the
conversion page, thank_you.html, and adds
one more conversion to the total for the
B-kitten-C combination this visitor saw.
Note that the Conversion Script will record
the conversion only if the user first came
through the test page, executing the Control
Script and the Tracking Script. Visitors
who happen to reach the conversion page
through some other route won’t be counted
as conversions.
Points to remember
• siteopt.js is not a static file. It is generated
dynamically based on the values passed to
it and on the specifics of your experiment.
• The Control Script also checks to see if
the page is using HTTP or HTTPS. If the
page is using HTTPS, it will make a secure
request for the siteopt.js file.
Points to remember:
• Any individual visitor will be counted only
once for visits and conversions on a given
test, no matter how many times he or she
visits the test or conversion page. This
makes each test an “apples to apples”
comparison of first-time visitors.
• GWO cookies have a lifetime of two
years. A conversion must happen within
two years of the visit to the test page to
be counted. (And if you finally get your
conversion 23 months after first contact
with the customer, congratulations!)