AB Tasty is a web application to run A/B tests easily. It requires no technical background and little IT involvement. Once the tag is installed on all pages of the site, marketings teams can design and set up their tests in a few minutes directly through a WYSIWYG editor and access results with bespoke KPIs easily through the report interface. A/B tests, split tests, and MVT are possible so e-marketers can confirm their conversion hypothesis and measure its impact on sign up rate, e-commerce conversions…
The ease of set up for such tests should not hide the fact that A/B testing requires solid foundations and a strict methodology to get results and eventually improve your conversion rate. The worst plan would be to run blind tests.
Here at AB Tasty, we do our best to make sure our clients get the most of our tool by encouraging them to adopt such a methodology. We advise them to run a complete diagnostic of their site to figure out what are their conversion killers. This diagnosis can refer to many different elements such as the web design, the UX, the value proposition, the copywriting… and requires many different tools to gather both qualitative and quantitative data (surveys, usability tests, web analytics, etc.).
For convenience purposes, most clients first look at their web analytics software to get such insights. Google Analytics, Site Catalyst, and many more tools are great for highlighting some behaviors, spotting shopping cart abandonment, or finding what pages of your funnel is the least efficient. But this data mainly covers hits, events, or page views and is in no way related to your users. They don’t give you a clear overview of what each user actually does on your site.
That’s where tools like KISSmetrics fits in. They are focused on users, not just page views. For instance, KISSmetrics can help you understand what actions a specific user or group of users actually take on your site. With information related to users, analysis possibilities are endless.
With KISSmetrics, you can easily create a segment of users – let’s say your topmost active users (the ones that connect every day to your site) and analyze their behavior: what are they doing on the site, what features do they use more than the average users and so on. This information can give you tremendous insight into your key features and eventually encourage you to promote them more heavily.
Similarly, you can compare any metrics related to the audience segments you created. For instance, you can compare sign up rates for your service for segment A (ex: users who used your live chat support) and segment B (users who didn’t). Discrepancies in results may indicate a correlation but not necessarily a causation. Typically, what you want to do with this information is to test it. You need to validate your hypothesis: people who actually use our live chat feature have a greater likelihood of signing up.
Integrate AB Tasty with KISSmetrics
To test this hypothesis, you can use tools such as AB Tasty. Results will be directly readable in the AB Tasty report interface, but let’s say you have already configured reports and funnels inside KISSmetrics, wouldn’t it be easier to segment your existing reports based on which test variation your users were exposed to?
This is something that is actually really easy to do with the new AB Tasty connector for KISSmetrics. Similar to the Google Analytics or Site Catalyst integration, you can link your A/B tests to your KISSmetrics accounts directly from the test options panel. Just go to the options panel and select the “Third party tools” option. In the popup window, just select “integrate” in front of the tools you want to use. A new option for KISSmetrics is now available.
This integration creates a new KISSmetrics property named “AB Tasty Test – 12345” (12345 being the test ID) whose value equals the variation name the user was exposed to. You can then use this property to create specific reports, or segment your existing funnels as described below.
The use of both KISSmetrics and AB Tasty can help you better understand your users, improve your segmentation, and figure out new test hypotheses and measure their impacts on your users easily.