It should come as no surprise that the Return on AI Institute, its expanding advisory firm, is in high demand five years later. After two hectic years of generative AI experimentation, companies are moving from experimentation to execution and optimising the business value of AI tools and applications. However, most businesses are still in the early stages of that learning curve.
Many are asking the same question that Srinivasan and Davenport discussed at that cafe: Why do the vast majority of businesses see their AI pilots fail while only a small percentage—5–10%, in Srinivasan’s estimation—can derive significant value from their investments in AI?
According to one conclusion drawn from the firm’s in-depth research and interviews with executives of more than 100 multinational corporations, the elite few didn’t only outsource AI to outside consultants and vendors. To create business value from AI, they instead developed internal talent and capabilities within their companies and implemented company-wide strategies to use the technology to solve business problems.
According to Srinivasan, “it’s not about how much data you have or how big a model you build.” It has to do with human conduct. How do you alter the organisation’s mindset and behaviour? This is how AI can be used to produce significant value.
The Human Element
The Roai Institute’s approach to working with organisations is driven by human transformation rather than technological transformation. There are two main pillars in the framework:
Preparedness for AI. They start by evaluating the organisation’s fundamental skills and AI readiness, including infrastructure and talent. Are culture, talent, and leadership all strategically aligned? Which models of return governance do they have? Do they have a reliable infrastructure for enterprise technology and data?
Increasing understanding of organisations: The institute develops a 12- to 18-month advisory program based on those insights to expedite progress. It consists of workshops, one-on-one coaching, and introductory sessions on AI.
According to Srinivasan, “the goal is forming consistent habits.” “To ensure that these behaviours persist at the individual and organisational level, we provide them with a foundation of knowledge about AI in addition to continuous simulations and social learning.”
The institute’s AI primer sessions have been attended by almost 1,000 business executives from the healthcare and life sciences, financial services, industrial manufacturing, and technology sectors over five years. According to Srinivasan, the people who have profited the most from AI ultimately end up with the highest return on investment. Additionally, they are prepared for the organisational and strategic adjustments that AI necessitates, not just receptive to them.
He claims that they already recognise the significance of AI. Although they are unsure of its precise nature, they are aware of its significance and fear falling behind. They possess what I refer to as a “transformation trigger,” which indicates that change is urgently needed. It’s really difficult without that. It makes no difference if your business is small, large, or even a leader in the field. You will turn into Kodak if you don’t change.
According to Srinivasan, people, not technology, are ultimately responsible for ROI. “I’ve known businesses that have made significant investments in data infrastructure and warehouses, and five years later they’re asking, ‘Where’s the beef? What happened to the return? They haven’t done the mental work, which is why they are stuck.
Specify results first, not last.
Companies’ AI investments have become more urgent due to the intense hype surrounding generative AI, competition, and direct pressure from CEOS or boards.
According to Srinivasan, “leaders and executives say AI is life or death for them, but even though CEOs say AI is important, I don’t see it getting the airtime and management focus it needs.”
Instead, too many CEOS delegate responsibility for establishing AI strategy and prioritising use cases to other departments within the company, allowing data science or AI teams to select pilots and pitch them to the company. Why? Srinivasan notes, “They simply don’t feel comfortable thinking in new ways because they don’t have the fluency in AI.”
Too many businesses make the crucial error that business use, not business strategy, is driven by the attraction of new technology.
“From my perspective, AI turns into solutions chasing problems,” he states. Because AI comes first and results follow, about 90% of AI projects fall short of expectations.
Business executives must pinpoint specific business issues that AI can resolve on its own to see a true return on their investment.
Srinivasan cites a big wireless carrier as an example of a successful business that faced a pressing issue with customer attrition. In one market, they tested an AI solution that decreased churn by 3–5%. Little potatoes? He points out that by implementing it in every market, the telecom company was able to reverse the decline in its customer base while also creating hundreds of millions of dollars in additional value.
According to Srinivasan, “it’s easy to come up with ideas and chase them because you get excited.” However, defining the problem necessitates speaking with customers, aligning research, and applying critical thinking. It is a challenging process. Businesses that make the error of prioritising AI are failing to clearly define their goals. Right now, a lot of time and money are being wasted on that with gen AI. Successful businesses say, “This is what AI should do for us.” Our North Star is this.
According to Srinivasan, businesses should take the following parallel routes when figuring out how to make sure their AI strategy maximises returns:
Deployment in a tactical manner: This entails figuring out how AI can enhance current procedures, like when an HR team uses Chatgpt to create job descriptions or when customer support representatives use a copilot to answer frequently asked questions. This approach ultimately increases revenue and cost savings, and it is where the majority of businesses today seek AI’s benefits.
Deployment strategy: Even though measuring the short-term value in terms of investment is more difficult, ambitious, and challenging, a second approach is just as important. Adopting a “digital native” mentality and investigating novel AI-powered business models to become a disruptor—a possible Uber or Airbnb in a crowded industry—is known as strategic deployment.
According to Davenport’s research, businesses that are fully committed to AI—that is, attempting to transform their operations, develop novel business strategies, and open up new markets and product lines—have a 2.7-fold increased chance of becoming more competitive in their respective industries.
Srinivasan acknowledges that it’s not simple.
According to him, “it requires long-term commitment, appropriate levels of investment, an experimentation mindset, more employees fluent in AI, and CEO and board level commitment.” “Your senior management group must put in double effort. They must be convinced. But according to Srinivasan, businesses must start both tactical and strategic initiatives if they want to reap the greatest benefits of AI in the near and distant future.
And what is his last piece of advice for executives who want to stay ahead of the ROI curve?
He advises, “Have a clear philosophy around how you’re going to measure value.” “Some projects are simple. You can demonstrate how AI is gradually adding value with and without it by running a transparent A/B test. However, in other situations, particularly when it comes to productivity and general artificial intelligence, it can be challenging to determine how much of the value comes from human decision-making versus AI.
He points out that successful businesses will already have a well-defined philosophy and procedures in place for calculating the return on investment of any new project or technology. They can and ought to use the same financial restraint when calculating AI’s return on investment.
He asserts that an algorithm is not necessary to determine the worth of an investment. “Culture has a greater role in measuring meaningful value than AI does.”
