Friday, June 7, 2013

Google Analytics Becomes A Robust Testing Platform With Content Experiments API

Google Analytics Becomes A Robust Testing Platform With Content Experiments API:
Our Analytics API enables marketers and developers to experiment and build new tools powered by Google Analytics. Over the past year, we’ve listened to your feedback and made improvements to the API such as expanding data points available and integrating with Apps Script. Our goal is to provide the most flexible and useful Analytics API on the web enabling you to do everything from build great apps to automate / expedite busywork. 


Today, we’re excited to share the launch of an API for Content Experiments — our tool for easily testing site content with programmatic optimization to achieve Analytics objectives. This API makes Google Analytics a full-blown A/B testing platform where developers of all types can leverage the power of Google Analytics to run their experiments. By utilizing our multi-armed bandit approach, you can maximize results by efficiently determining which assets on your site perform best to offer an improved experience for users. Multi-armed bandit experiments are powerful and efficient tools and with the new Content Experiments API, you can get even more from them.


The Content Experiments API allows you to pick and choose from all the testing functionality Google Analytics has to offer and to combine it into powerful solutions that best fit your particular needs:

Testing changes to content without redirects. 
The original Content Experiments JavaScript snippet made testing a breeze. To keep things simple and consistent for all publishers, the snippet causes a page redirect which may take away from the end user experience in certain cases. Now, with the new Content Experiments API, testing changes to content without redirects is both possible and easy to implement.

Testing items server-side such as the result set of a database query.
Major testing platforms typically offer changes on the client-side but not server side. With Content Experiments API you can now run tests on the server side and try things like implementing different recommendation or search algorithms to determine what works best for your site.

Testing with your own variation selection logic and use Google Analytics for reporting.
While the multi-armed bandit approach to experimentation is one of Content Experiments most powerful features, there are times where publishers and developers would prefer to decide for themselves how to serve variations - be it evenly or using proprietary logic. The Content Experiments API makes it possible for you to bypass our programmatic optimization while allowing you to continue to enjoy the powerful experiment reporting Google Analytics provides.  

Testing in non-web environments using measurement protocol.
For example, if you have a kiosk in your physical location (such as airline terminal or retail store) you can test different layout variations of content and features and determine what users can complete quickest or at highest value.

Developers are already putting the Content Experiments API to work and we’ve been hearing great feedback. Paras Chopra, Founder & CEO of  Visual Website Optimizer reports:


"We're thrilled about the possibilities opening up with the new Content Experiments API. This new API is specially designed to infuse the powers of Google Analytics into testing and experimentation domain. We're very proud to be one of the beta-testers with Google and soon we will start rolling out the integration of Visual Website Optimizer with Google Content Experiments across our joint customer base. When Google releases an API, it's a big move for the A/B testing industry and we're excited to be their launch partners."

Learn how to get started with our Content Experiments API on our developer site or if you’re still new to the platform, get an overview of Content Experiments in our help center.

Happy testing & experimentation! 

Posted by Russell Ketchum, Google Analytics Team


DIGITAL JUICE

No comments:

Post a Comment

Thank's!