For our fourth project Ray Lou and I were temporary hired to work for General Assembly and our task at hand was to conduct user research for a new online circuits class in Data Analysis.
We started out by mapping out students to target based on a preliminary survey our clients had conducted. We identified prospective students and reached out to them with a small incentive, hoping that they would give us an interview in return. We interviewed 10 people by phone for about 30 minutes each. After transcribing the interviews we analyzed the responses and broke them down into segments. As patterns started to emerge, we identified 4 personas. We created empathy maps for the personas and based on our initial assessments and research, provided recommendations as to how our clients should proceed with further research and course planning. You can view our presentations below as well as read our high level recommendations below.
General Summary, “common denominators”
Everyone knows the basics about analytics tools available
- They want to be guided and have confidence in GA suggestions
- Marketers are looking for a more automated method to analyze data
- Data Visualization is the most important to learn (among all industries).
- They all have a common base knowledge of excel
- Pivot tables are considered more advance
- Some want to increase their proficiency in excel
- Experts want to move away from it to more open sourced platforms
- Strong emphasis on Google Analytics for Marketers/UX
- Mostly website and social media metrics
- Time to dedicate to class: range from 2-15 hours
- Mentorship model would be very beneficial (8/10 agree)
- Workload can not be too heavy
- Important to be measured and get feedback
- Certificate should be granted (1 person indicated this)
- Use it to get better at their jobs (almost unanimous)
- Success two-fold measured: - be able to apply new skills to job immediately/ be able to communicate better with analysts and get overall view of what data analysis is
- both contributes to the work place directly as a result of taking this course
- Would like to learn one tool really well vs. get a general overview and resources to learn something really well (2 side to the coin)
- Interactivity is important - not question/ answers or pure information dump (they can find information online themselves, they pay to get more)
- they want to tell their stories with data (both marketing and data/consultant)
- Program can be gamified and tell a story throughout (1 person indicated this)
- Curriculum should be applicable today - business relevant, not purely theoretical. Not reading books, interactive way of learning.
Subjects seem to have a thorough understanding that they are not using data analysis to its full potential. They feel like they are missing out on valuable insights that could make them better at their jobs. Subjects want GA to educate them on which tools they should be using and how.
They want to get better at Google Analytics and they place a heavy emphasis on user-focused research (Marketing)
Important concepts and tools to cover: Google Analytics, Excel (basic and advanced), SQL, Data Visualization tools.
From our research we found that prospective students currently deal with data in some form on a weekly basis and want to become better at it so that they can improve on their job. The majority of subjects are over 30 years and well established in their careers. Success from the program is measured by being able to quickly apply the skills directly to their current work, so it is important that the curriculum consists of real-world problems, with a focus on user-based technology products. As they engage in the learning process they want to see incremental progress to stay onboarded.
It is important that the program has crystal clear milestones and steps of completion throughout the duration of the course. Interactivity is important, and being able to conduct most of the work directly on the web without needing to download additional software (like dash). Video tutorials should be super simple.
- Skipping tutorials/modules should be an option due to the varying level of skill for the different data tools. Three methods are described below:
- One way to gauge skill level is to require an assessment test before enrollment. Based on their proficiency in the proposed topics, students can receive varying levels of pre-work and/or choose to move past beginner modules of the class if they already know the topic.
- Develop two or more levels per module (Basic/Advanced). Students will take an assessment test before each new module and only get access to the module that fits their knowledge and skill level. This way some people get advanced tutorials while others get more basic.
- Build a wide variety of modules around content. Request users to answer questions upon enrollment to understand what their needs are and what tools are most important for them that they learn. After assessing, offer ten modules relevant to their personal learning objectives and needs, and hide the others. This allows for a broader reach immediately and a more narrow scope in terms of finding right mentors.
- Continuous motivation and feedback is important. The mentor will help with both feedback and questions. Users reported that releasing all the modules and curriculum at once will be overwhelming and demotivating. Some find the topic intimidating and want to move slow. General consensus that they are exposed to piecemeal understanding of data as a whole.
- Google Analytics is sought after. If not intended for the program, this should be marketed very clearly. Most subjects have taught themselves Google Analytics through video tutorials and Google search. It may be beneficial to incorporate certain functions like optimizing different methods to visualize data within Google Analytics.
- Data visualization should be the prime focus. A majority of the subjects create presentations frequently and want to better communicate data to their less technical stakeholders through visualization. The goal is to be better storytellers and convey the message across effectively.
Due to the online nature of the course, users want to learn on their own time and with a reasonable weekly time commitment. There should be a way to access both peers and mentors for quick feedback (something akin a forum).
Our user subjects with marketing/UX backgrounds are very excited about the program. Many have had positive experiences with GA before and are confident that their instructors will introduce them to the leading industry tools. Some are not sure what they want to learn and want a broader and general understanding. Across the board, students named specific areas of importance:
- Understand the data they work with daily
- Visualize data more effectively
- Work more efficiently “smarter”
- Derive more insight from data
- Gain proficiency to do the work without relying on other teams
- Deal with larger sets of data
- Understand customer segments and what questions to ask
- Create and test hypotheses
Next Steps - User Research:
Unanswered questions are:
Are the 10 subjects interviewed representative?
Survey prospective students and include questions regarding demographics such as age, gender, location, industry, level, motivation
Many subjects seem keen on increasing proficiency in Google Analytics. Figure out whether this is representative of all previously surveyed subjects. If so, find out if they are still interested in taking the course if it does not include Google Analytics. Make sure it is clearly communicated that Google Analytics is not part of the course.
Nail down breadth and depth in terms of course development. Create separate landing pages and A/B test which tools are more popular with your potential applicants. Send out different emails marketing the course two separate ways. Advertise certain tools in one email and a different set of tools in another ( basic tools such as excel and sql be standards) and measure which tools get the highest turnout.
Initial assumptions was that recent graduates would be interested in this course. However, our respondents did not fit this demographic. It is hard to tell whether this age group just is less likely to respond to requests for interviews or whether the target audience is older than anticipated. In terms of re-surveying or additional user research this subset of younger users is worth figuring out (as they obviously are an attractive demographic to have).