A/B and Multivariate Testing with Personyze is Simple, Yet Powerful

Personyze allows you to take multivariate testing your site’s experience to the next level; rather than simply testing across all users on your site, you can achieve a greater precision, avoid data pollution, and reduce necessary testing time, by testing individual audience segments you have targeted in each campaign. A/B and multivariate testing are built directly into the action creation widget in Personyze, so the process couldn’t be simpler:

  • Segment your audience, and create the changes to be tested

  • Drag and drop the individual changes into testing groups (with optional control group)

  • Select how the changes will be distributed to visitors within the customer segment (using a simple drop-down menu, and slider)

  • Lean back, and watch the data pour in

All other visitors to the website continue to see the standard version of the site, and this minimizes any lost conversions due to the testing. Additionally, you can set Personyze to delete the under-performing actions automatically, once it becomes clear which action is having the best results.


Additional Benefits of A/B Testing in Personyze:


  • Detailed statistical data to compare groups

  • Targeting based on CRM data

  • Unlimited custom goals and tracking

  • Campaign statistics are sent to Google Analytics, by default

  • Reporting to Google Analytics can also be turned off if you choose

  • Control group option, exclude a segment of visitors from all campaigns if you want

  • Use different methods of rotation to get exactly the data you want


Personyze robust testing capabilities allow you to analyze and optimize overall visitor experience on multiple pages, page templates, and external channels, such as emails or Facebook.  By rotating page versions, Personalized content, popups, and banners, wherever your personalized changes appear, they can be included in a single test.


Data-Driven A/B Testing Has Never Been Easier

  • The ability to test the effects of versions on limited customer segments, rather than all visitors to your site, saves you time by eliminating irrelevant data, and also reduces wasted conversion, by not showing the action to customers who don’t match the segment that it is intended to market to.

  • You can create a control group of any size in your multivariate test; this allows you to ensure that some of the users will not see the changes you are testing, and also allows you to extract more accurate data from tests where personalized changes overlap across your audiences.
  • You have the ability to test a single action across multiple pages, page templates, and external channels. This means that an action can be embedded or tied to many locations on your site(s), but tested as a single variable.

  • Personyze allows you to see the performance of each group according to city, country, and other variables, and also contains simple KPI metrics by default, such as clicks and impressions, in addition to whatever custom KPI indicators you wish to create.

Marketing Scientifically Doesn’t Have to Be Hard

We all know the value of data-driven marketing, and using an empirical approach doesn’t have to be difficult. With Personyze, you can see the data on changes you make in real time, using a simple interface, and with auto-pilot optimization, can even set the campaigns to automatically delete when they under-perform in the multivariate test. At Personyze, we’ve done everything we can to make this process easy and intuitive.

A/B Testing Your Campaigns in Personyze is Simple

It’s hard to imagine how the actual process of setting up a multivariate test could be any simpler than it is, in Personyze. While your audience segmentation and campaign actions can be as complex as you want them to be, turning that campaign into an A/B or multivariate test is as simple as pressing a button above your actions, to turn on A/B Testing Mode.


Screen Shot 2017-02-13 at 4.22.53 PM

Once in A/B Testing Mode, you drag and drop individual actions (representing changes or additions to your site) into the different testing groups, adding as many groups as you need.

Once you have grouped the actions appropriately, there are only two essential A/ B testing settings to be adjusted. One is a drop-down menu, which decides how to show different A/ B testing groups to different visitors (such as randomly every time a visitor opens the page, or only one version per visitor, etc.).

The other is a simple slider, which determines the percentages of campaign audience members who will see these multivariate test changes. This allows you to show more of the audience one group of actions, or another.

When this process is complete, you can set a control group if you want, using the control group option from the drop-down menu.

 Once these settings are saved, all you have to do is wait for the traffic to flow, and the data to be gathered. If you want to set the campaigns to automatically “self-destruct” when they are outperformed, you can set a timer for the multivariate test; when the time is up, winner takes all, and the under-performing campaign is automatically deactivated.

With a process as simple as this, there’s no obstacle to prevent you from employing data-driven, empirical marketing in your website personalization, without any other 3rd party testing required.