A test measures page performance based on the standards set by project type:
baseline, split, or multivariate. You can run multiple tests on the same page, in the same
project, over time or simultaneously.
From the navigation menu, click Laboratory.
Double-click a multivariate project from the list and then click
Enter a name and description. Leave the test
state as Pending until you are ready to verify
the test in a browser or share the test with site visitors.
(optional) Select Collect and store standard data to
record data about your visitors while the test runs. Standard data includes IP
address, browser type and version, operating system and other information.
(optional) Select Specific audience segment to limit
this test to a particular group of visitors. For example, you can limit a test
to visitors have a certain operating system or web browser. Then click
Browse to choose from a list of available segments.
Determine the percentage of page traffic (Traffic Throttling) to be presented
with testing or targeting content. Any remaining traffic is presented with
default content. It is often useful to lower throttling on high traffic pages in
order to limit consumption of Optimize Events. To change the current value,
click the percentage and enter or select a new number. 0.1 is the minimum. Traffic throttling can be
changed at any time.
Note: By default, Optimize reevaluates returning visitors to the test or target
page as new visitors and uses the current throttling percentage to
determines what type of content to present. To present returning visitors
with the same content they viewed in previous visits, clear the
Reset project throttling cookie check box.
(for fractional factorial tests) Choose the level
sets you want to include in the test. A level represents one piece of
content; a level set represents a grouping of levels. The list of available
level sets is based on the number of page factors that were entered when the
project was created. To test two headlines, two buttons, and four images, you
would select the 2 2 4 level set.
You can use projected test duration to help you choose a level set. Enter your
current known conversion rate and visits per day (available from baseline test
results) and select a Conversion Rate Lift: Low, Medium, High, Custom. Optimize
then calculates the test duration in number of days for each level set.
(for fractional factorial tests)Drag each level set to the appropriate factor
for the test plan. For example, to test two headlines, drag a level set 2 to the
factor that represents headlines. If you selected a level set that has an equal
number of levels for each factor (2-2-2), Optimize automatically assigns the
After you have assigned levels, add web content for each factor in the list.
Click the blue + icon for a factor to open the
content input box.
In the Add Content window, enter a level name
(optional) Select the Control check box to
designate the level as the standard against which other levels will be
Note: If one factor contains a control level, then the other factors in
the test must also have control level.
Manually enter the test content for the level or click
Browse for existing content to select
(optional) Select the Use current page content
check box to use the current HTML page content for this level. Here you
can also add a script to the current page content that runs each time
Optimize presents the current page content as part of an experiment.
Note: Optimize does not rotate test content when current page content is
used in an experiment.
Click Save to commit the content to the level. A
check mark appears on the level icon to indicate that it is complete.
Repeat the add content steps for each level in each factor.
code that you want to run before the page is fully rendered in a browser. The
code you enter here runs before segmentation, content streaming, or split page
code that you want to run after the page is fully rendered in a browser.
(optional) Click Custom Data Collection to collect data
such as user IDs, product codes, or shopping cart totals from your test and
conversion pages and use it later in custom reporting.
This option requires you to set up a custom data table and configure your
test page. See Custom Data
for more information.