Artificial Intelligence Adoption a Matter of Wait-and-See vs. Active Exploration
In response to IBM’s “AI in Action” survey, Shobhit Varshney, Senior Partner and VP with IBM Consulting, talks about the differences between “Leaders” and “Learners” when it comes to artificial intelligence
As the artificial intelligence (AI) revolution continues, it’s become clear that there are two distinct categories of AI users: organizations that are actively exploring how to best use the new technology and those on the sidelines, waiting to see how AI develops before taking the leap.
Which approach is best? In partnership with The Harris Poll, a recent IBM survey of 2,000 companies across the U.S., U.K., India, Japan and Germany found that enterprises with a proactive AI strategy, defined as being guided by an “AI Leader,” see a 25% boost in revenue growth. Another 27-38% of Leaders report that their AI program has led to improvements in five critical areas: productivity, cybersecurity, customer experience, marketing effectiveness and streamlined processes.
In detailing the survey results, the “AI in Action 2024” report says that other companies with a wait-and-see plan led by “Learners” may be avoiding the expenses and mistakes made by early adopters of AI solutions, but could also be missing out on the benefits.
Are You a Leader or Learner?
Does industry and location determine how a company approaches AI? Interestingly, the survey showed that Leaders came from a wide variety of industries. The most (18%) were from retail, while finance and manufacturing were just behind, at 16%, and those from telecommunications companies numbered 13%. Geographically, the largest percentage of Leaders were from the U.S. and India, while 7% were based in Japan.
In addition, the survey reveals three essential traits that separate Leaders from Learners. “It’s the ability to align the organization, harness the power of data and move quickly into production,” says Shobhit Varshney, Senior Partner and VP with IBM Consulting, where he leads the AI, Generative AI and IoT business across the U.S., Canada and Latin America. “The great thing is that these are traits anyone can adopt.”
Perhaps the best quality in AI Leaders is their skill at getting various departments focused on AI objectives. “This enables the AI Leader to create a strategy that considers business objectives, internal technical capabilities as well as the organization’s ability to manage change,” says Varshney.
Varshney also points to their ability to see the value of data and prioritizing the establishment of a data infrastructure. This, he says, provides a foundation that’s essential for successful AI solutions.
Another quality that separates Leaders from Learners is their appetite for innovation. “They’re willing to move quickly from ‘trying’ to ‘doing,’” says Varshney. “Leaders aren’t content to linger in the pilot phase. They shift from pilot to production as soon as reasonably possible and they have the discipline to iterate and pursue incremental improvements.”
The Best Use Cases
According to the survey, Leaders prioritized four specific use cases where they wanted to see what their AI solutions could do: customer experience, IT operations and automation, virtual assistants and cybersecurity.
All of these cases are clearly tied to business outcomes, from driving top-line growth, improving productivity and efficiency and shielding the company from risk. Up to 80% of Leaders feel these are worth an investment in AI technology to see what could be achieved.
Choosing a use case where you can take the first steps in an AI solution can take some thought. It’s natural to assume that it should be tried on a simple, easy-to-automate area, but that may be the wrong approach. Looking for areas where AI could make the most difference may provide a deeper, more informative view of the technology.
“Ultimately, where to begin experimenting with AI is unique to each business,” says Varshney. “Look at where your organization has areas that can be improved, these could be initial areas to explore.”
Creating an AI Road Map
A realistic, actionable roadmap is a key part of a Leader’s toolbox, with 85% of those surveyed saying they used this as their guide toward optimizing their solution. The majority of Leaders list four dimensions to their roadmaps that address their unique needs, capabilities and IT investment. According to the “AI in Action” report, an AI roadmap must, in general, have the following components:
- A strategy centered around a holistic vision for the broad application of AI across the organization that will be embraced by both the business and IT leadership
- An investment in toolkits that simplify the development and implementation of AI apps, are supported by internal staff and, if necessary, supplemented with third-party expertise
- A focus on data management that ensures AI initiatives are grounded in accessibility and governance—this also supports the need for customized AI solutions
- A broad array of targeted application use cases, with a priority on those that will apply AI where it’s most impactful
The Possible Roadblocks for Learners
Why are Learners taking the backseat as their competitors move forward with AI? One big reason is a breakdown in communication. “Beyond building consensus across the company leadership, it’s important for leadership to recognize the importance of celebrating wins and communicating authentically to gain—and maintain—buy-in,” says Varshney. “Leaders with strong storytelling skills can generate excitement and buy-in by connecting the dots between the theoretical benefits of AI and the practical benefits.”
A troubling element of the survey, according to IBM’s report, is that just 11% of Learners “believe in their ability to access and effectively manage their organization’s data to support AI initiatives.”
With data management such an important element in AI solutions, this is obviously a big hurdle to overcome. “Approximately 80% of respondents highlight worries about data security, data privacy, data quality and numerous data silos within the organization,” says Varshney. “As data environments have become more complex—multitudes of data silos, new data types, etc.—it’s clear that Learners don’t have absolute confidence in the ability of their data infrastructure to properly serve their AI needs.”
Finally, another roadblock is how some organizations will conduct multiple proofs of concept on various solutions and become frustrated with the process, shelving the solution until some undetermined time.
“It’s important to try, test and iterate,” says Varshey. “But Learners who can value incremental progress rather than holding out for the perfect ‘big win’ are those who will continue moving forward from piloting to broad implementation. Test, improve, but always keep moving forward is the way to make progress.”
How to Become a Leader
Despite the survey’s message that Leaders are pulling ahead of their Learner friends in AI, there’s still time for Learners to catch up and surpass the competition. “Just get started is the best advice,” says Varshey.
“Start by assessing where you are, where you want to be, and—this is important—don’t be afraid to ask for help. See where your organization has gaps. It may be that in addition to investing in technology you need to invest in your people. Remember that no one is an absolute expert at this point, so just jump in.”
And while being stuck in the Learner category may not be the worst position to be in, you’ll ultimately want to make the effort to turn your organization into a Leader, says Varshey.
“Two-thirds of Leaders report seeing a greater than 25% improvement in revenue growth rates since the start of their AI journeys,” he notes. “The good news is that the report lays out a clear roadmap to success for the Learners category, enabling these organizations to accelerate their efforts while minimizing challenges to success.”