That’s the question being asked in boardrooms around the world, and higher education is no exception.  And given disruptions in education funding, answering this question has never been more urgent, with postsecondary institutions grappling with how to maintain service levels to students, faculty, and staff. 

Higher education leaders must come together across technology and mission areas to mount an effective response by: 

1. Grounding in strategic value for higher education 
2. Understanding how to harness the power of AI 
3. Aligning the institution to a high-performing data strategy 

While Slalom is an experienced implementer for a range of cloud, data, and AI platforms, this data and AI strategy approach is tech-agnostic by design.  In a fast-changing landscape of tools and solutions, the goal of this article is to provide you with a durable North Star. 

To borrow the words of Simon Sinek, "Let’s start with why."

When it comes to data and AI strategy for higher education, let’s start with the “strategy” part. 

By starting with “why,” you can avoid getting lost in the details of “how.”  And your fundamental reasons for investing in data and AI all come down to strategic value. 

At the highest level, strategic value in higher education is of two types. 

• Type 1 strategic value is achieved by enhancing the inflow of resources into the institution, including both monetary and nonmonetary resources.  Major categories of resource inflows include public funding, tuition and fees, grants, gifts, and partnerships of all kinds.
• Type 2 strategic value is achieved by enhancing the positive impact when those resources are expended.  Examples include improvements to mission areas of student recruitment and success, faculty experience and research, administrative and business functions, and engagement with alumni and local/global communities. 

Collectively, Type 1 and Type 2 strategic values give us a map of all the desirable destinations for our data and AI journey.  And this map helps to align our teams to ultimate outcomes, which range from boosting alumni fundraising to reducing dropout rates. 

Which leads to the next question: How can data and AI get us to these destinations? 

AI and related technologies represent a powerful engine for creating strategic value.  And you can expect a lot from a good engine.  You can expect it to be efficient, to perform, and to do so reliably.  But if you want to travel from point A to point B, that engine will need to be connected to a steering wheel, some tires, and a gas pump! 

In this analogy, functional expertise is the steering wheel, integrations to your systems of execution are where the rubber meets the road, and the necessary fuel is high-performing data. Based on Slalom’s experience in demonstrating the strategic value of AI solutions, we strongly recommend bringing these components together to optimize your odds of success.

• Examples of functional expertise include mission areas such as enrollment management, academic advising, student success, and career services, as well as administrative areas such as IT, finance, and HR

• Examples of systems of execution include student information systems, learning management systems, and service platforms

• Examples of data that can be made more high-performing include information about campaigns, applicants, students, faculty and staff, alumni and fans, and partners

Keep in mind that these holistic requirements for ensuring your data and AI investments yield the desired outcomes apply equally to machine learning, expert systems, generative AI, and the coming wave of AI agents.

Now that we’ve defined the key components needed to harness the engine, let’s turn to the vital question that data strategy answers: How to put your data to work in service of strategic value?

No doubt, your institution is already getting value from your data.  And there is an opportunity to create even greater value by leveraging data to solve the problems that were intractable just yesterday.  That process begins with data strategy. 

We define “data strategy” as the starting point for a process that includes execution and enablement.  Data strategy is where you establish vision and goals, develop a plan with engagement from key stakeholders (including priority use cases), and analyze current state gaps to create a roadmap to the future state.  Executing that data strategy may involve implementing new technology, or maybe you just need to get more out of the tools you already have.  And this technical execution should be run in coordination with supporting people and process enablement. 

This strategy/execution/enablement process is iterative, not a single journey.  You don’t define a data strategy, then execute and enable it one time.   

Your institution likely already has data strategy at both an enterprise level and various departmental and campus-unit levels.  However, few colleges and universities have solved the larger puzzle of how to bring data together across domains to connect the dots and unleash the full potential of AI solutions.  And this type of ambitious vision is necessary to provide a consistent (and competitive) experience to applicants, students, faculty, researchers, and other key constituents. 

So how can you align technologists, data practitioners, academics, and outcome owners to a holistic data strategy that will drive the next wave of innovation?  Once again, start with why.

Even for full-time data and AI professionals, it can be a challenge to keep up with the new products and features being launched all the time.  Which is why we developed the Slalom Framework for High-Performing Data. Think of this framework as a ladder with five rungs, where each rung represents a durable goal, regardless of what specific solutions are used to achieve them.

Traditional data strategy is rightly focused on the foundational layer of trust.  This is where people, process, and technology are deployed to ensure that your data is available, scalable, and governed for the purposes of security, privacy, and compliance.  This is the rung of the ladder where you identify your data sources, understand where that data is stored, and manage how that data is accessed.   

This trust layer is concerned with how to make data available at scale in a modern architecture, avoiding unnecessary migrations and preventing adverse impacts to performance.  It is in this trust layer that we ask: Do you have appropriate identity and access control, encryption, and privacy management?  And do you have the people and process enablement needed for your future state, including robust policies and change management? 

Next is insight.  This is the layer that connects your data to human understanding and includes business intelligence and analytics tools.  Human understanding is supported on this rung by machine learning, predictive algorithms, and generative AI.   

Any combination of people, process, and technology that assists in transforming raw information into knowledge, trends, patterns, and hypotheses can be mapped to this portion of the framework.  This includes use cases that feature AI-assistance in simplifying complex information environments into human-friendly summaries, answers, and recommendations. 

From insight, you ascend to the next layer of decision and action.  On this rung, data and AI-driven outputs are leveraged to support human action and/or the new breed of agentic AI that can take independent action. 

We predict that in the coming months data-driven decisions and actions will be parsed along a continuum of risk and repetition, where AI agents will be used with increasing frequency to execute simple and routine actions.  This evolution will reclaim human time for thinking through strategic decisions, being proactive, handling complex escalations, and making human connections. 

After decision and action, you get to the final boss levels of accountability and strategic values, which are all about measuring, managing, and ultimately achieving Type 1 and Type 2 strategic value.  These layers involve test design, examining leading and lagging indicators, comparing relative benefits at the portfolio level, and making adjustments as needed. 

Taken together, this model for high-performing data provides a framework for mapping a myriad of specific data and AI solutions to a ladder of goals, with each tier building on the next. 

Finally, here are three key points to remember about data and AI strategy for higher education:

• First, start with why.  The “how” of data and AI is changing by the day, but the strategic goals for AI and the framework for high-performing data give you a durable North Star as you navigate a sea of acronyms, providers, and tools. 
• Second, data is the fuel for the AI engine.  AI is an engine that needs functional expertise to guide it, systems of execution to create impact, and trusted data that can power your organization’s journey to strategic value.
• Third, don’t let perfection be the enemy of progress.  Moving iteratively from data strategy through execution and enablement requires a pragmatic approach.  Use the data that you already have, even as more data and better data become available, knowing that your data estate will never be fully complete.  No matter.  Keep going. 

The good news is data strategy does not mean making your data flawless.  Your imperfect data, which is basically all data, can be made more high-performing.  The data that you have can be used to create a trusted foundation that supports insight and enables action.  These data-driven decisions and actions can, in turn, provide measurable results that ultimately create strategic value. 

If you’d like to discuss your institution’s AI or data strategy, please reach out to jennie.wong@slalom.com or kris.karaikudi@slalom.com to schedule a call.