[ADD] website: AB Testing: Unpublished file to be tranformed in rst

This commit is contained in:
Frédéric Gilson 2016-09-21 13:25:06 +02:00 committed by Yannick Tivisse
parent 83f1caf9a1
commit ae5a121877

View File

@ -0,0 +1,87 @@
=====================
How to do A/B Testing
=====================
<p class="text-muted">
A/B testing is the process of testing two versions of a page to see which one performs better in reaching your business objectives.
</p>
<h4>Here is a basic guide for your first A/B testing:</h4>
<div class="o_planner_paragraphs">
<div class="row">
<div class="col-sm-1 col-md-2 icon"><span class="fa fa-bullseye"/></div>
<div class="col-sm-11 col-md-10">
<p><strong>1. Define what to test</strong><br/>
The main goal is increasing your ROI. Figure out the best page(s) to
test. Here are a few examples of good testing variables:
</p>
<li>Call-to-actions (i.e. buttons) wording, size, color and placement,</li>
<li>Headline or product description,</li>
<li>Forms length and types of fields,</li>
<li>Product pricing and promotional offers,</li>
<li>Images on landing and product pages,</li>
<li>Amount of text on the page (short vs. long).</li>
<br/>
</div>
</div>
<div class="row">
<div class="col-sm-1 col-md-2 icon"><span class="fa fa-files-o"/></div>
<div class="col-sm-11 col-md-10">
<p><strong>2. Create the page variation</strong><br/>
</p>
</div>
</div>
<div class="row">
<div class="col-sm-1 col-md-2 icon"><span class="fa fa-bar-chart-o"/></div>
<div class="col-sm-11 col-md-10">
<p><strong><a href="https://www.odoo.com/documentation/user/9.0/website/google_analytics.html" target="_blank">3. Configure a Google Analytics account</a></strong><br/>
This is necessary in order to record the statistics of your test.
</p>
</div>
</div>
<div class="row">
<div class="col-sm-1 col-md-2 icon"><span data-icon="&#xe085;"/></div>
<div class="col-sm-11 col-md-10">
<p><strong>4. Create your campaign in Google Analytics</strong><br/>
Simply follow the wizard by creating a new experiment in the <strong>Behavior >
Experiments</strong> menu of <a href="http://www.google.com/analytics/" target="_blank">Google Analytics.</a>
To learn more, take a look at their <a href="https://support.google.com/analytics/answer/1745147?hl=en&amp;;ref_topic=1745207&amp;;rd=1" target="_blank">online help</a>.
</p>
</div>
</div>
</div>
<table class="table">
<thead>
<th><h3><span class="fa fa-thumbs-o-up"/> <strong> Good practices</strong></h3></th>
<th><h3><span class="fa fa-thumbs-o-down"/> <strong> Bad practices</strong></h3></th>
</thead>
<tbody>
<tr>
<td><strong>Know how long to run a test before giving up</strong><br/>
Don't give up too early! Wait for your test results to be significant. Take a look at the statistical
confidence in Google Analytics; It should be at least 95%.<br/>
Don't give up too late neither because poorly performing variations could cost you revenues.</td>
<td><strong>Don't test versions at different time periods</strong><br/>If you
test one version one week and the second the next one, your results won't be
accurate, or worse, wrong!</td>
</tr>
<tr>
<td><strong>Do many A/B tests</strong><br/>
Don't despair if your first A/B test turns out to be a lemon. The key in optimizing conversion rates is to do a ton of A/B in order to put together all the most effective configurations.</td>
<td><strong>Dont surprise regular visitors</strong><br/>
If you are testing a core part of your website, include only new visitors in
the test. You want to avoid shocking regular visitors, especially because the
variations may not ultimately be implemented.</td>
</tr>
<tr>
<td><strong>Test only one variable</strong><br/>Try only one variable at a time, otherwise you won't be able to clearly interpret the results of your modifications.</td>
</tr>
</tbody>
</table>
<div class="well well-lg">
<strong>Next actions:</strong>
<li><a t-att-href="prepare_backend_url('website.action_website_configuration')" target="blank"> Install A/B Testing</a></li>
<li><a href="https://www.odoo.com/documentation/user/9.0/website.html" target="blank"> Read the documentation about A/B Testing</a></li>
</div>
</t>