Initial findings from the digital conversion study

Feb 1, 2016

In July, the Station Analytics team started an on-going investigation into the long-term site interaction trends that correlate with higher levels of site engagement and donation. This project uses Google’s BigQuery tool to load granular Google Analytics data starting from April of 2014 and building into the future. This tool allows us to analyze user data across longer periods of time than what is currently available in the Google Analytics interface and to perform more complicated custom calculations.

The longer timeframe and customization allow us to answer more complex questions like:

What is a user’s most frequent source?
What type of page does this user usually enter on?
How many stories does a given story reader see in a session?
What can we learn from looking at recency and frequency?
How do donors use station websites differently from other users?  

This project is important because it helps us to focus in the immediate future, and in years to come, on using site analytics data to improve in many areas:

Build better products to address the needs and interests of our users.
Create membership messaging that speaks to our individual users in ways that resonate and drive donation.
Promote content in ways that increase engagement for our existing users and grow our new audience.

Between May 2014 and July 2015 we collected data on almost 195k individual transactions from 85 different member stations participating in the Station Analytics Service. From this dataset we were able to learn some interesting things about users who have donated online.  

Sixty-six percent of these users gave in their first visit.  Some of this speaks to how we set up and use our websites - since donation often happens on a different property we sometimes have difficulty connecting the data with the use of our main web property, but it also indicates that some donors may chose to transact online even though they are not users of our websites. This points to an opportunity to tell web-savvy donors about our digital offerings.

A few other key takeaways:

  • Online donors come from traffic sources that reflect their loyalty (direct, search) but seem less likely to respond to social.
  • Online donors stream more but read less frequently. However, when they do read, they consume more than the average user.
  • Support takes time. Online donors are likely to visit support related pages in the visit before they donate.

To learn more, you can download the slides and view the recording of the August webinar on nprstations.org.