Student Analytics: Focusing on a Critical Issue as a Catalyst for Creating a Data Culture

Institutions have been using terms such as “data driven” and “data informed” with regard to decision making for decades, but, as Tambellini research noted in a recent blog post, the issue of poor data hygiene impedes any meaningful progress for many institutions. So, while many aspire to create a data-informed culture, significantly fewer have the data governance, data architecture, and expertise to create a trusted data environment. But, perhaps despite this, there is a pathway forward that can accelerate progress in achieving this desired state of data utopia.

A combination of factors has combined to create a dip in student enrollment that only promises to worsen in the coming years. It is essential that institutions gain much better insights into enrollment trends and student success metrics, including retention and persistence rates, and develop appropriate strategies to successfully mitigate these challenges, which undoubtedly will be coming to an institution near you in the not-too-distant future.

I will not attempt to argue that institutions that have implemented robust data governance, eliminated data silos, and built cross-institutional support are in a significantly better position than those that haven’t to take on the challenges facing higher education. However, the availability of vibrant student analytics platforms and associated services may provide a vehicle to overcome these shortfalls.

In the recent Tambellini Practice Guide: Leveraging Student Analytics, I covered many of the features and strengths of student analytics solutions and best practices in use today─primarily student analytics vendors offering a SaaS + Services model. Typically, the solutions are built on a cloud platform such as Amazon, Google, or Microsoft. Data is ingested from the data sources to build the data model needed for the specific application. Data from the student, learning management, and constituent relationship management systems are typically among the sources used in building the models. Several vendors provide robust professional services, such as data engineers and data scientists. Some offer accelerators, primarily through prebuilt data models, that significantly reduce implementation time, and results can often be achieved in a matter of months.

Focusing on a single issue of vital importance to the institution can greatly accelerate achieving the vision of becoming a more data-driven organization. Identifying a specific issue, such as increasing enrollment yields or improving student retention, can help marshal the institutional support needed to drive such an effort. As my colleague noted in the recent blog post The Data Problem in Higher Ed, maturing “data management is not an IT function; it is an institutional imperative.” This includes the following components:

• Data Unification: Siloed data will not produce robust analysis. Identifying data sources and the gatekeepers of that information is critical to implementing an analytics initiative. Strong data governance can help break down these barriers and create ongoing cooperative practices.
• Establishing Multidepartment Support: Using data from multiple departments and systems requires cooperation from the entire campus. Having support from executive leadership will help overcome some of the difficult challenges inherent in this work. This means bringing information technology, institutional research, and functional areas to the table.
• Identifying Security and Privacy Policies. It’s important to be transparent with campus constituents about what data will be used and how. Creating security and privacy policies ensures this knowledge can be found and understood easily.

Taking on a specific issue can drive each of these aspects, particularly when senior leadership is involved. As data sources are identified, ownership issues can be addressed, a data management plan can be implemented, and data governance can be hardwired into the solution. The SaaS + Services model can be very attractive to schools that need access to the resources necessary to build and manage these analytics initiatives. The services offered can vary from a limited engagement, assisting with a particular issue, or building and managing the entire solution. Other offerings include a subscription for a specified number of hours per month. These services grant institutions access to data engineers and high-level analysis and models without the burden and cost of maintaining those resources internally. The vendor shares the cost among its multiple customers on the platform.

More vendors are offering sophisticated predictive analytics models using machine learning, natural language processing, and other customized AI algorithms to provide personalized, in-depth insights to drive institutional decision making. This strategy provides institutions with an option for overcoming poor data hygiene and accelerating the journey to becoming a data-driven environment. 

Takeaways:

• While a comprehensive data management program is essential for long-term success, focusing on a specific issue can serve as a catalyst for this effort. The SaaS + Services solution can provide data governance support as part of the particular solution.
• Moving from the abstract to a specific, tangible solution can help jump-start the longer-term effort needed to build a data-informed culture. Though some data models can be built quickly some in less than six months a data management effort will require sustained effort and cooperation among departments. Using the initial project as an accelerator can create a sense of urgency that may be difficult to muster otherwise.
• Building a successful data culture will require a cross-functional focused effort. Avoiding shortcuts that circumvent data governance efforts will likely bring success in the longer term and increase the likelihood of creating a data-driven institution.