The conversation started simply enough, “We’d like for our patients to be able to schedule appointments through their Amazon Echo device.” I kept listening as the client explained, “We just want to make it easier for our patients. This is a great way for us to use artificial intelligence technologies; our executive team wants us to move fast.” I paused and then asked several questions, trying to understand the problem and why they decided to start with Alexa. The short answer was that smart devices like Google Home and Alexa makes artificial intelligence (AI) look easy, hiding the complexity of a broader business transformation that needs to take place in order for AI to deliver ROI.

We’ve heard from our clients that artificial intelligence can feel overwhelming and that they want to be able to see immediate results. But AI is a journey and needs to be understood as such. It’s important to think strategically about your entire company and the role that AI will play in its future.

1. The ability to make sense of unstructured data – texts, images, videos, audio, drawings, etc. which were not possible with previous technology
2. Learning from historical data
3. The ability to interact with the users in a human way – speech, chat, video, etc.

When building an AI strategy, it needs to be grounded in these areas and focused on how an enterprise can use AI to solve earlier unsolvable problems. AI is all about taking advantage of your data to learn and improve from past performances, as well as using it to deliver new types of experiences. In order to be successful with an AI effort, you need to start from your strengths to transform and improve what you are already doing today. It’s about being honest with yourself regarding your organization’s readiness and thinking big but starting small.

Most companies begin their AI journey in pockets and silos across the organization. Maybe it’s within finance to have more accurate forecasting. Or in marketing to better understand the highest value customers and how to better target and reduce churn. Or quite possibly your recruiting software is using AI to help you find and match the right candidates. Running AI in silos, while it can be highly successful and bring significant ROI, it eventually needs to tie to an overarching, companywide AI strategy.

Your strategy needs to involve both the AI you buy off the shelf embedded in the software products your organization runs today and the custom AI algorithms that your employees build internally to help run your business., You also need to decide where you plan to invest as your own differentiator and where you’ll just pick up the net benefit of the AI that is already delivered to your employees via Microsoft in Office365.

Research from the Center for Information Systems Research (CISR) at MIT Sloan School of Management shows that data-driven organizations need to do more than just hire data scientists, but also build enterprise-wide capabilities to scale the practice across the company. Companies with strong AI strategies know that collaboration is a foundational factor in driving success. Collaboration ensures that AI is baked into the business, rather than sprinkled on top, to paraphrase Stitch Fix CEO Katrina Lake.

So, what are the components of an AI strategy, especially if you’ve already got numerous projects underway? You need to think about:

1. AI across your Business: What areas/business functions/processes are critical that differentiate your business where you need to invest and build custom models? Where do you want to utilize AI in third-party software? Think broadly across every business unit.
2. AI Talent: Do you have the talent that understands how to leverage AI technology and is that talent aligned and deployed to the right places in your organization. Can your teams look at an existing business problem and ensure that they are considering AI functionality that is ready to scale today? What is your automation and people strategy? As much as AI is about data, it is more about people.
3. AI Data and Technology: Do you have the projects underway to ensure you’ve got your data supply chain set up in the right way? Are you taking a realistic view of AI technology to leverage a 2-3 year time scale? This is about taking stock of where you are at and ensuring you have the foundation in place to get there.
4. AI at Scale: As with most significant technology transformations, organization change management becomes very essential, to scale the changes. Think about creating a center of excellence that integrates all the initiatives and owns the AI agenda for the enterprise.
5. AI Responsibility: Does your organization’s ethics agenda cover technologies and innovations? Do you have a data ethics impact capability? Which business processes using AI have human interventions at the key decision points? It’s critical to build customer and employee trust by taking action on digital ethics.

It is never too late to start thinking about forming or updating your AI strategy. Like the Alexa example from earlier, jumping right is in good, but you need to know where you are going. AI is not a magic ingredient, as it is complicated, and you need to understand what you are doing and why. AI is a powerful set of technologies that you need to start figuring out how it will help to transform your business.