Overview

Gaia is a cross platform streaming media network focused on conscious media. The website hosts over 7000 videos ranging from yoga to conspiracy theory and has hundreds of thousands of monthly subscribers. Between 2015 and 2018, Gaia’s monthly subscriber base grew from 100K to 500K, largely accomplished via acquisition testing, marketing and paid media. To sustain the success from rapid growth, we wanted to focus on engagement.

We created a feature team to focus on AB testing to increase engagement. From a qualitative perspective, we wanted to give members a fulfilling, delightful video viewing experience. From a quantitative perspective, we focused on the following goals:

  1. Increase median minutes watched

  2. Decrease time to 1st and 2nd video view for new members

  3. Ultimately, reduce 1 month churn by 5%

Methodology

Test rigorously: Every experiment was an A/B test, measured by a proprietary AB testing software, built in house. 

Consider qualitative: Validate and explore results by talking to customers. We conducted moderated and unmoderated task analysis, sent surveys , reviewed screen recordings and held in depth interviews.

Fuel experiment ideas: Leverage learnings from strategic research, customer feedback, and from a broad range of internal company members.

 
 
 
 

Experiment: Onboarding

While Gaia offers video content in many formats: yoga, documentaries, short films, episodic content, ect; primarily new members expect to either “practice” or “watch”. Currently, onboarding options offer an array of topics and video formats, and a user’s onboarding selections result in a homepage filled with personalized recommendations. By simplifying the onboarding options to “practice” or “watch” a user ends up with personalized recommendations that are more likely to fit their expectations.

 
 
 

Control: New users can select multiple interests

 

Test: New users have two options, ‘watch’ or ‘practice’

 
 

Results: Test wins!
By simplifying onboarding we decreased time to 1st video view for new users.

 
 
 
 

Experiment: Global nav redesign

We wanted to test changing the interaction paradigm and copy for Gaia’s global navigation. The control nav was a drop down pattern and we wanted to test always exposing the sub menu to get users to relevant content faster. Additionally, the control nav included terms like ‘Seeking truth’ and ‘Transformation’ which were ambiguous for new members. Eliminating those terms and surfacing lower level topics (deep space, crystal healing, holistic nutrition, ect.) would help new members find videos.

 
 
 

Qualitative layer

In addition to AB testing, a task analysis was conducted with 10 test users that are interested in conscious media topics, but have never logged into gaia.com. 5 people were shown the control, 10 were shown the test variant.

 

Results: Test wins!
• For new users on desktop screen sizes, we decreased time to video playback
• Other tenure cohorts not impacted negatively
• Other screen sizes not impacted negatively
• Qualitative findings including an unmoderated task analysis showed a more successful experience

 
 
 
 

Experiment: Topics ecosystem

We wanted to test more effectively tagging the site and surfacing ‘lower level’ topics on a landing page and throughout the experience. This would allow new users to explore threads of interest and increase engagement.

 
 
 
 

Results: Test wins!
• In the test group, 9% of members who visited gaia.com also visited the topics page.
• Small lift in median minutes viewed compared to control
• Small lift in time to video playback compared to control
• ‘Topic’ pages like ‘Pineal Gland’ or ‘Plant Medicine’ increased acquisition by topic