Prerequisites for an ROI-led data and AI strategy

As organizations pursue next-generation data and artificial intelligence (AI) technologies to improve business outcomes, what’s getting in their way?

Some of the key challenges we see are determining the right level of investment (across people, processes and technology), accessing the right talent, managing hybrid data sets and architectures, and integrating systems and data—both external and internal.

Organizations are at different stages of their AI journeys. Some are developing roadmaps while others are validating existing ones. Some are focused on what I call the “core lift”— moving out of the data center and into a modern cloud data stack. Others are seeking to raise their AI maturity level to execute transformative efforts.

Across all stages, success depends on having roadmaps and business cases that are actionable, practical, and ROI-led. Success also requires getting the organization into a more agile mindset of delivering value in days and weeks versus months and years.

In this blog, I share an approach to developing a data and AI business case and roadmap that’s ROI-led and supported by responsible practices. I also discuss two frameworks we use to help clients target challenges that are best solved with AI solutions and that will add meaningful value. Finally, I offer tips for how to keep the ball moving forward.

Creating a compelling business case

Next-gen data and AI solution delivery requires a next-level business case. Here is a simplified example: We will invest $50 million in data and AI (representing the total cost of ownership across technology, in-house talent and external consulting services) to unlock $150 million in value over 5 years.

AI business cases should articulate specific ROI and get everyone on the same page. This requires collaborative workshops with executive leadership and even employees in the field. It also requires defining an AI use case inventory or solution catalog and demonstrating how that will deliver innovation, reduce costs, and/or boost productivity.

Building an actionable roadmap requires understanding where you are today and defining where you want to go. Equally important, however, is identifying what it will take to get there. What architectures, tools, security, governance, methodologies, and use case inventories will support a compelling business case around data and AI?

Assessing AI use case feasibility and ROI

Answering this question requires a comprehensive feasibility assessment of use cases and their opportunities to increase savings, efficiency and quality. That’s why we developed a “preflight checklist” that helps identify the numerous prerequisites and dependencies for deploying next-gen data and AI solutions. We like to say, “there’s a lot of ‘ditch digging’ to do.”

If you have the investment to pursue an AI initiative in your business case, what are the steps you need to take to proceed? Can you take these steps and deliver? Are you blocked by anything, such as the lack of a data set or source system integration or being stuck in a data center? For example, cloud is industrialized at this point and a must have to support next-gen data and AI.

It’s important to demonstrate that a use case has a business context and will have a definable improvement outcome with a clear and quantifiable ROI.

There also are what we call midflight considerations. Maybe an organization has a roadmap but hasn’t launched it yet because they are wrestling with what investment to make, what they need to do and when, or how to address security and compliance. Maybe they did a preflight checklist that identified risks or limitations, such as needing to source the right talent or ensure the architecture is in place at kickoff. Maybe they’ve stalled or are not going fast enough. When people ask ─ “Are we done yet?” ─ no one knows if they are a month away, a year away, or five years away, as “done” hasn't really been defined.

A key to addressing these issues is to ensure there is an abundance of clarity around business case outcomes and ROI for solutions in the roadmap. At the same time, organizations should maintain a level of flexibility and creativity knowing that data and AI initiatives will evolve as the initiatives kick-off and / or accelerate. We generally call this a “starting with the end in mind” approach that strikes a healthy balance between alignment of expectations for the end solution(s) and accommodating the latitude to pivot as priorities change with business needs.

Driving value faster with a more agile mindset

Delivering next-gen data and AI solutions and earning that next level of investment requires demonstrating value rapidly and often. It also requires training and upskilling people to move away from old methods and mindsets and to think differently. It’s about showing progress quickly and having the right people in the room.

Once you have your use case inventory, proven AI delivery frameworks help rationalize, prioritize, and explore high value use cases. The aim is to build early pilots and minimal viable products for evaluation, proofs of value and prototypes in a very rapid and iterative way, demonstrating progress and receiving feedback. Using an agile methodology designed for data and AI engagements. Then, you can move to operationalize and industrialize these solutions and ultimately progress to a build, integrate and deploy scenario. The organization begins to drive value based on the business case, expectations, and investments made.

The key is instilling a more agile and iterative mentality—delivering outcomes in days or weeks versus months and years.

Some organizations are executing their roadmaps but need help climbing what we call the “AI summit.” This involves standing up an effective AI organization. We believe the four imperatives for doing this are to envision, explore, engineer and expand AI solutions.

This agile mindset also involves the rapid experimentation of new use cases and AI solutions and building momentum to get the investment to operationalize the solutions.

Ultimately, you want an environment where everyone's operating in a more intelligent and automated way and driving significant productivity, with more focus on solutions and less focus on the underlying technology and complexity. This is the North Star everyone's trying to reach.

You don’t have to figure it all out before you start

As you consider how your organization can become more iterative and agile and show value quickly, you also need to be okay with exploration of opportunity and be willing to abandon use cases that have less value. It’s about having the right mindset and processes. Our point of view is that you don't have to have everything figured out all at once.

Organizations may struggle to “do it all” by themselves or feel the need to address every initiative in the business case, and then they get bogged down in the analysis. However, it’s important to decide who will own the initiatives, as well as what will get done and when. Will you address these initiatives today or at some point in the future? Do you have the skills in-house, or do you need to bring in a trusted partner?

These are the kinds of questions I enjoy helping our clients answer every day. Please connect with me to learn more about ROI-led AI business cases and roadmaps.