BI Trends

Predicting the Business Value of Artificial Intelligence

By Sriram Parthasarathy
Share on LinkedIn Tweet about this on Twitter Share on Facebook

Over the next five years, 91 percent of business leaders expect to see new business value from Artificial Intelligence (AI) implementations, according to a new report from the MIT Sloan Management Review in partnership with BCG Henderson Institute. The report, called “Artificial Intelligence in Business Gets Real: Pioneering Companies Aim for AI at Scale,” was based on a survey of over 3,000 business executives, managers, and analysts from organizations around the world. It’s a long report but worth the read. In particular I noticed a few key highlights:

>> Related Content: 2018 State of Embedded Analytics Report <<

1. More and more businesses will start incorporating predictive insights into their applications as quickly as possible. The fact that 91 percent of respondents expect to see new business value from AI implementations indicates the value of predictive analytics is going up. End users who are demanding actionable foresight will benefit by consuming predictive insights directly in their everyday application workflows. Applications that embrace AI fast will become viral, and the rest at some point fall off the earth.

2. Pioneers are continuing to invest and expand their commitments in AI, scaling the technology throughout their enterprises. In the past, many AI success stories came from isolated problems tackled by knowledgeable AI managers. Moving forward, the impact of AI will scale to more roles and a broad diversity of skill levels.

AI can play a critical role in many problems as it scales. On a manufacturing floor, AI can identify which machines are at risk of failure in the next two weeks and trigger predictive maintenance. In an accounts payable department, AI can tag invoices for late payments and spend leakages. For a sales organization, AI can help predict which customers will churn and recommend a preventive campaign to keep them.

The number of problems AI can tackle is enormous—but the critical constraint has always been access to resources and talent. Data science talent is hard to find and expensive to keep. Plus, data scientists are not always aware of the internal expertise the current team has gained about internal data, use cases, and customer insight over the years. In that context, smart data science tools and platforms that help an application team create predictive insights without needing to write R or Python code are tremendously valuable. These smart wizard-driven predictive tools make it easy for software teams to create and distribute predictive insights as part of the existing application workflows quickly, allowing business users to get future insights and take action inside their application.

For AI to scale, solutions need to enable everyone on an application team to create and distribute insights. As more business users can create and deploy predictive analytics, it will help companies react quickly to the demands of the AI-powered economy in the future. Application team feels empowered and motivated to learn new things, and that helps companies retain good talent.

3. Pioneers prioritize revenue-generating applications over cost-saving ones. AI can be applied to increase revenue or reduce cost. In the past more focus has been on lowering costs, but the MIT study shows that soon, emphasis will shift to revenue-generating applications. Some examples include predicting whether a particular customer will buy a certain product, click on a product link, or subscribe for a service.

4. AI creates both fear and hope among workers. Some people express fear that AI will ultimately take jobs away from people. That is a negative way to think about it. AI may impact some of the jobs that are easy to automate, but in most cases AI will act as a partner to the human operator. AI can help people in all industries do their jobs better, increasing productivity and improving decision-making capabilities. In that context, future workers need to learn to use the insights provided to them by AI applications as part of their work lives.

5. Management support and unclear business value are the top AI roadblocks. According to the report, the most common problems with AI projects continue to be “lack of management support” and “unclear business case.” These go hand in hand: Any successful product manager knows how to create prototypes and gather feedback from customers, then use those to get attention from upper management. Building a clear business case can help move an AI project along.

6. Will “AI Centers of Excellence” become popular? When the business intelligence (BI) revolution started, many enterprises set up a “BI center of excellence” as part of the executive offices. I have heard successful examples of this as well as a few horror stories about the impact. In general, having a central place to disseminate AI best practices, give guidance, and incubate projects is a great way to cross-pollinate learnings from projects across a single company.

My biggest takeaway from the MIT report is that companies are beginning to use AI to drive business value. Application teams in particular have an opportunity to find new ways to incorporate AI, machine learning, and predictive insights into their application. Time to market is going to be very critical in their success.


Originally published January 17, 2019; updated on July 31st, 2020

About the Author

Sriram Parthasarathy is the Senior Director of Predictive Analytics at Logi Analytics. Prior to working at Logi, Sriram was a practicing data scientist, implementing and advising companies in healthcare and financial services for their use of Predictive Analytics. Prior to that, Sriram was with MicroStrategy for over a decade, where he led and launched several product modules/offerings to the market.