Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
Assessing data readiness for AI is a critical yet often misunderstood step towards adopting artificial intelligence. Recent headlines like CIO’s “AI Data Readiness: C-Suite Fantasy, Big IT Problem” illustrate the murkiness businesses face:
In my 12 years of helping businesses build their data strategy, I’ve learned executives need to take ownership of assessing their data readiness for AI. To start, ownership of data has migrated out from IT towards business teams (where it should be) thanks to factors like the explosion of cloud technologies, low-code and no-code platforms, and an increasingly tech-literate workforce.
As a result, IT has shifted from the traditional keeper of data and technology to more of a coaching, guiding, and governance-facilitating role. By helping teams like sales, finance, and operations structure their data, IT enables teams to integrate data into their own strategic decision-making.
This shift has led business leaders from around the organization to take more control and accountability for their systems, technologies, and data. Slowly dying are the days of saying, “That’s IT’s problem.”
Now that data is more tied to business functions than ever before — and that leveraging data demands more cross-functional effort than ever before — executives from around the organization should work together to assess their data readiness for AI. Ideally, CEOs and CDOs should lead the charge, working closely with other executives atop each business function.
Data readiness is the level of alignment across your data, architecture, technologies, processes, culture, and skill sets to achieve business goals. That’s the textbook definition. But that answer doesn’t mean much until we ask a few key questions so we can go from theory to practice.
Start by asking, “Readiness for what?” And readiness for AI isn’t enough of an answer. Instead, think in terms of what challenges you’re trying to solve. AI is a broad spectrum. If you want to build a chatbot that can answer employees’ questions about company policies, you’re probably ready right now. All you need from a data readiness perspective is a few PDFs.
But let’s say you have more advanced AI ambitions like:
Note the above ambitions don’t mean you need a full-blown AI strategy to begin your data readiness assessment. But they do illustrate that you need a vision for AI, one clear enough so you can ask, “What high-value AI use cases can we identify across the organization based on our vision and business objectives?”
This approach sharpens your focus on how sophisticated of a data ecosystem you’ll need. It’ll also keep you from 1) trying to boil the ocean and 2) focusing on too narrow of a use case, two common traps businesses fall into when assessing their data readiness for AI.
Now let’s walk through what assessing data readiness looks like. We’ll expand on what it means to create a vision and identify use cases, then use those to guide an initial high-level assessment that reveals where your gaps are.
Establishing a clear vision for data and AI sets you up for success in the following steps. Think of your vision statement as the compass and your use cases as the map.
Your vision should consist of short, clear, and actionable statements that give your business a north star for its data and AI goals. A few sentences should suffice. The goal is to present a vision that other executives and senior leaders can easily understand and translate into action.
Here are some example vision statements representing three different industries:
Notice how each vision includes both a business-facing statement (i.e., focus on business goals) and a user-facing statement (i.e., focus on who will use the data and how). This informs the data operating model you’ll need to build by showing you how data needs to flow through your organization.
With a clear vision established, you can begin identifying the potential use cases your data will need to support. A quick AI use case workshop is a high-impact way to come up with a set of use cases that are both feasible and will deliver ROI sooner than later.
Let’s use the telecom vision statement from step one, “We want to hyper-personalize user experiences so our customers feel treated like people, not commodities. Our frontline employees will use data to deliver exceptional experiences at every touchpoint in the customer journey.” This vision might generate AI use cases like:
We can elaborate this to other industries, too. Healthcare providers, for instance, could identify use cases across their clinical practices, operations, claims processing, remote patient monitoring, and data management.
Note the goal here isn’t to commit to every AI use case you come up with. Rather, it’s to find the handful of priority use cases that will drive the focus of your data readiness assessment.
The goal of an initial high-level assessment is to:
Your high-level assessment will focus on the five key pillars detailed in the graphic below.
Remember the goal isn’t to boil the ocean here (i.e., try to assess everything at once). Instead, identify and prioritize which pillars need the most attention, then use that to roadmap your assessment. Likewise, your priority use cases help point the way, as do tools like a data readiness questionnaire.
Say your initial use cases cluster around marketing. Your assessment should zoom into the data, technologies, skill sets, and key performance indicators (KPIs) supporting the marketing department. Or if customer experience (CX) emerges as a priority area, then you’ll zoom into the data sources and technologies supporting your loyalty program, sentiment analysis, contact center, and other CX functions.
Note it’s helpful to bring in a third party as they’ll give you a fresh perspective. They’ll ask questions that internal stakeholders are prone to take for granted, and draw insights from helping other clients succeed. That outside perspective will help you rapidly establish baselines, begin creating an AI and data strategy, and start putting AI and data governance frameworks in place.
In my experience, what many executives discover during their initial assessment is that different parts of the business are at different levels of data maturity. And that’s good. This shows you the areas of strength you can lean into, creating momentum for strong collaboration across the organization.
The gaps uncovered in your initial assessment show you where to prioritize your focus. Start by assessing which strength or weakness you want to address first. Remember your deep dives should tie back to your priority AI use cases. Time spent fixing issues that don’t map to one of these use cases will waste time and effort (i.e., don’t boil the ocean). Tackle each gap with tools like:
Let’s say data management issues emerge as a significant gap. You could investigate this by analyzing data accuracy rates across different systems, examining data entry processes, and/or evaluating your data cleansing.
Next, develop a remediation plan for each gap. These plans should outline specific, actionable steps to address each issue, taking into account required resources, timelines, and key milestones. And make sure the right people are involved to get buy-in and alignment (e.g., if unlocking data silos between sales and marketing, bring together roles like your CFO, CRO, and CMO).
Last, prioritize improvements based on strategic impact, weighing factors such as:
Note not all gaps need to be addressed at once. Rather, focus on the fixes that will most impact on your AI readiness, enable priority use cases, and build a foundation for the future. A strategic approach like this will help the more mature parts of your organization implement AI safely and effectively as less mature departments to catch up.
Ultimately, the goal of data readiness is to prepare your data, teams, and technologies to realize the AI use cases that will propel your business. But data readiness isn’t an easy concept for everyone to understand, which makes getting buy-in a challenge.
In my experience, culture is a consistent issue. There’s often varying levels of internal resistance when it comes to data readiness, especially for AI. When people don’t understand the goals of data readiness, it provokes fears like their performance is under a microscope, or that AI is coming for their job. This plays out in ways like:
The result is poor communication and collaboration, which keeps critical issues from surfacing (e.g., discovering your departments use inconsistent data definitions).
But you can get ahead of this resistance with consistent communication. For example, help everyone understand the goals of readying your data for AI with messaging that proactively addresses their concerns. Keep communicating that message with regular progress updates. Reinforce your goals, the first use cases you’re developing, and the impact they’ll make.
Moreover, when talking to specific teams (e.g., HR, product development, operations) emphasize what they stand to gain through data readiness:
And as you lead the assessment of your data, remember not to point fingers or call anyone’s baby ugly. Instead, focus on finding opportunities to make everyone’s lives easier. That intent will support a more forthcoming team around you who’ll buy into your vision.
Only 12% of organizations report their data is ready for AI according to Precisely, opening up a competitive advantage to businesses who move early. But readying data for AI is complex work. The time it takes varies, too, depending on the current state of your data and the eagerness of your organization for AI.
Whatever your stage, it’s highly recommended to choose an AI consulting partner with strong expertise and a proven track record in data readiness. At WillowTree, we help companies at every stage of their data and AI journeys, from end-to-end strategy and implementation to targeted services that help you break through roadblocks.
Learn more about how we can accelerate your journey by exploring our Data & AI consulting services.
One email, once a month.
AI-fueled experience transformation.
Smarter products. Faster builds.
Automated marketing and operations.
Proudly a telus digital company
