django-gwo v0.2 documentation

Management Commands

pull_from_gwo

Updates the data with the data from Google Website Optimizer. Local versions are deleted and recreated.

$ ./manage.py pull_from_gwo
Purging local experiments
Pulling experiments from Google Website Optimizer
 Building experiment Example Experiment
  Pulling sections
  Building section Top Nav
   Pulling variations
   Building variation Original
   Building variation Not Showing
  Building section Bottom Nav
   Pulling variations
   Building variation Original
   Building variation Showing
Done.

generate_variation

Prints to the console a template with sections replaced with the variation specified in the combination. This makes it easier to recreate the template(s) after the experiment is finished.

This management command needs the template name and the combination number:

$ ./manage.py generate_variation blog/post_detail.html 5

and you can put the output to a file by:

$ ./manage.py generate_variation blog/post_detail.html 5 > post_detail.html

generate_variation then requests the combination from Google Website Optimizer. The combination specifies which variation to use with each section. For example, if you had a section in your template specified as:

{% set_experiment "Example Experiment" %}
{{ gwo_experiment.control_script|safe }}
  <h2>{{ object.title }}</h2>
  {% gwo_start_section "Top Nav" %}<p class="other_posts">{% gwo_end_section "Top Nav" %}

depending on the combination specified, the results could look like:

<h2>{{ object.title }}</h2>
<p class="other_posts" style="display: None">

The {% set_experiment %} tag and {{ gwo_experiment }} variables are gone, the {% gwo_start_section %} and {% gwo_stop_section %} tags are gone, and the contents between them is the value of the variation.

Note

Included templates are not touched.

If your experiment included markup on templates that were included with the {% include %} tag, this command does not look at them.

Note

track_click tags are not removed.

The track_click scripts will work fine without an experiment running. It also saves on adding them back in if a follow-up experiment is run.