Earlier this year, we shared a blog post about our methods for surveying our bilingual email subscribers. We wanted to learn more about our users, their needs and preferences, to serve them better. The data analysis happened in two phases. In phase one, we focused on a high level summary of over 9,600 survey responses. And in the second phase, we explored areas that warranted a deeper analysis, such as how different subtopics interacted and the main characteristics of those who subscribe to our business list.
Phase I - High Level Summary Phase I offered a snapshot of our USAGov email subscribers. We learned that the respondents generally:
- Prefer to receive emails weekly
- Are 55+
- Use USAGov emails to help themselves
From the survey responses, our vocal subscribers seem relatively established in their lives. Most aren’t interested in housing topics, like shopping for a mortgage or landlord complaints, maybe because 45% already own homes. Another 45% identified as retirees, so information about running a business or finding a job might not be of interest to this group. 11% are parents of school-aged children, so those respondents may not be concerned about education information for young children. Overall, 83% of those who responded to our survey have attended college, and almost half had earned at least a bachelor’s degree. These inferences from the survey responses allow us to shape how we present our email content and find ways to communicate to the people identified beyond an email address.
Phase II - a Deeper Dive The high level analysis ultimately made us curious to learn more about specific areas and dive into subsets of information from the questions asked. To best engage with our email subscribers, we sought out to learn more about:
- Business-related email content and those subscribers
- Topic overlaps - cross-sections within our subject areas, such as health, benefits, and education information.
- Overlap of roles - the instances when a respondent identified in multiple roles, such as a veteran who is also a homeowner and teacher.
For example, in 2015, we inherited a subscriber list from Business.USA.gov, when the website shut down. This survey was our first chance to learn more about this subset of followers. We’d made the assumption that most business management subscribers own businesses. But from the survey, we found that only 30% of business management subscribers own a business, most of which have been operational for at least one year. Now that we know, we can better address the needs of prospective and actual owners, share content to help nascent businesses stay afloat and aspiring owners understand business loans or steps to starting a company.
Additionally almost 90% of our subscribers want information about multiple topics. The most common combinations include government benefits, disasters and emergencies, and health.
Similarly, we learned that 75% of subscribers identify with multiple roles. A key consideration from our analysis of overlapping roles is the need to balance audience size and interest. For example, 575 respondents identified as both parents and homeowners. In our dataset, those 575 people comprise over 53% of parents, but only 13% of homeowners. Scenarios like these remind us to leverage these role overlaps to share government information where it will yield the strongest impact, based on our reach and engagement goals.
What’s Next? We’re excited to use these analyses to fuel our content strategy and better serve our email subscribers and agency partners. We’ll use these findings to adjust how we frame topics, based on our users’ preferences. These results challenged some assumptions we’ve held about our subscribers. Now we have data to ground changes, like splitting some subscriber lists and clarifying our topic buckets.
From a partnership perspective, these findings will allow us to bring more insights to conversations with other agencies. The information from the email surveying exercise has value beyond the “walls” of USAGov. To that end, we’ll look for opportunities to share these findings with others, both internal and external to GSA.
