By using website you agree to our use of cookies as described in our cookie policy. Learn More

TDWI Upside - Where Data Means Business

Successful AI Demands Careful Planning

Before you can exploit the benefits of AI-powered analytics, you must build a firm foundation that informs your AI strategy.

Nobody would build a house without a good set of detailed blueprints showing the completed structure and the components needed to put it all together. It's no different for a corporate artificial intelligence (AI) project.

For Further Reading:

When You're Ready for AI But Your Company Isn't

AI Techniques You Need to Know

Putting AI to Work Protecting Your Data

However, even as organizations are trying to determine how to use the power of AI to benefit their businesses, many are still missing the foundational planning necessary to develop, deploy, and maintain successful AI programs. To ensure your AI initiative meets your business needs, build a good foundation that establishes clear goals and sets up a solid plan.

Step 1: Build Your AI Foundation

Before embarking on an analytics strategy for an AI project, assess your organization's AI baseline by answering these three key questions:

  • How is AI defined at your organization?

  • Why is your organization using AI?

  • What business value do you expect from AI?

These questions might seem very simple, but they are surprisingly hard for many organizations to answer. In my experience, most organizations don't take the time to build their AI foundation, which causes their AI initiatives to struggle or completely fail.

Let's explore each of these questions. 

How Is AI Defined?

I recently led a workshop in which participants were asked what AI meant to them. I heard as many definitions of AI as there were people in the workshop. Several viewed AI as robotics, some viewed AI as math, others viewed AI as artificial life -- like Rosie the robot or Data from Star Trek -- and a few viewed it as augmented intelligence.

With such disparate meanings of AI, it is essential to establish how AI is defined at your organization. Document it and share it with all employees to give everyone a common understanding when they talk about AI and the organization's AI initiatives and goals.

Why Use AI?

Hopefully your organization is not deploying AI just because it's trendy. To have a successful AI program, you must determine how you expect AI to affect your business. I find that many organizations cannot articulate what they want to do with AI, which means they can't explain the value that it can bring to them or their customers.

A good place to start to answer "Why AI?" is to identify how AI analytics can help reduce a pain point for your business and/or customers. Then you can align this goal with a strategic corporate initiative to provide a road map for AI.

For example, Rogers Communication, a large Canadian telecom, wanted to enhance customer satisfaction and preserve its leadership in Canada's media and telecommunication sector. Rogers' goal was to gain a better understanding of each customer's need and to be as precise as possible in information presented to each customer. Rogers developed an AI program to do just that. By combining data and AI, Rogers can now successfully predict where a customer is in their buying journey and thus provide their customers with the information they seek.

What Business Value Do You Expect?

This is the most critical question in the planning process. The last foundation block is clearly defining the business value expected from AI. Document the business value so leaders can easily determine the success or failure of the AI initiatives.

In addition, quantifying the expected return on investment will help justify the project. From the example above, Rogers' Communication realized a 53 percent reduction in customer complaints, which qualifies and quantifies that their AI program was successful.

Step 2: Create a Strategy

Once you've established your AI foundation, it's time to move on to developing your data strategy for AI. This is not the same as identifying how to generate data. AI initiatives rarely suffer from too little data. Rather, organizations have trouble accessing and making sense of their data. A data strategy for AI solves that problem by providing a document that outlines needed data, access policies, the condition the data should be in, and the timeline for the needed data.

A data strategy:

  • Ensures the data can be used, shared, and moved easily and efficiently
  • Provides rules on who has access to which data
  • Views and manages data as a valued corporate asset
  • Establishes common terminology, methods, practices, and processes to manage and share the data in a repeatable manner

Understanding how the data is going to be used is imperative to ensure the correct data is wrangled, cleansed, and aligned to support the needs of AI. Your data strategy must establish, manage, and communicate to end users your policies, procedures, and mechanisms for effective data usage. Include a plan for storing the data so people across the organization can easily and quickly access what they need and, at the same time, clearly address the security around sensitive data, including access rights and how data will be protected.

In addition, analytics involving AI must be monitored to ensure accuracy and that the business objectives and values are realized. If you can't show business value, then your AI project is failing. In that case, go back to your AI foundation and revisit your structure and strategy. Perhaps you need another internal conversation to determine if you are ready for AI and whether your organization has the data and analytics expertise to support it.

A Final Word

Because data is a key element of AI, a good strategy built upon a solid foundation will ensure that the data for AI is properly stored, packaged, integrated, and governed. Having clear goals and success metrics for AI supported by reliable, trusted, and well-curated data will ensure that your AI projects are effective and meet your organization's goals.

About the Author

Kim Kaluba is a senior product marketing manager in data management at SAS. She has 20 years of experience in data management, including sales, marketing, and enablement. Kaluba received her business degree in marketing and management from Stetson University.

TDWI Membership

Accelerate Your Projects,
and Your Career

TDWI Members have access to exclusive research reports, publications, communities and training.

Individual, Student, and Team memberships available.