The history of American enterprise is littered with examples of technology that never got off the ground. Despite all the promise business leaders ascribe to new technologies, organizations end up not adopting the new software or hardware that—on paper, at least—seemed revolutionary.
You don’t have to look far for examples. The “why” is harder to determine.
Clearly, there are instances of failed tech rollouts no one could have predicted, but these are few and far between. More often, implementations of new technology fall flat due to a failure to manage factors other than the technology itself.
In my experience as CEO of an artificial intelligence-powered platform for wealth management and insurance firms, the tech—provided it functions as specified and does so efficiently—is essential. But equally important, if not more, are the people and process; in other words, how willing and amenable is the workforce to changing behavior to leverage the new technology, and how smoothly can it integrate into the organization’s operations?
Organizations embarking on new technology implementations would do well to dig deep on these questions, particularly when it comes to the adoption of AI platforms and solutions.
Behavioral change is more difficult for some workers relative to others, and in any organization, there will be individuals who immediately jump to use a new tool or adopt a new approach to doing their jobs. On the other end of the spectrum, there are workers for whom change of any sort is difficult; they are biased to the status quo.
Technology leaders must address both camps.
AI provides a great lens through which to view this challenge. Early adopters of AI technology in an organization typically have a sanguine view of its capabilities—it’s the new thing, and it’ll change the world. It has an almost magical quality for these individuals. On the other side, the most recalcitrant workers typically view AI as a threat, suspicious of what they see as an untested technology solely aimed at replacing human workers.
The reality is that the most accurate perception is somewhere in between. AI technology holds the potential to complete mission-critical tasks much as a human would, helping organizations cut costs and scale up more quickly. Yet it often doesn’t do this immediately, rather, an algorithm must learn how to do its job, and it can take time to achieve its potential and garner full return on the investment.
For behavioral change to be sustained and new tech implementation to have staying power, tech leaders must both manage the expectations of the early adopters and simultaneously make the case to the holdouts of the value of AI, both to them individually and the organization. For example, rather than replacing them, AI technology, given time to learn, can help workers do their jobs more easily and free their time to do more valuable work, and, by the way, it can make a company more profitable and help it grow faster.
It boils down to being straight with the workforce on what AI can do and what it can’t. It’s crucial for tech leaders to get the balance right, tamp down the lofty expectations of early adopters and buck up the dimmer perspective of the laggards. By meeting or exceeding the expectations of both groups, tech leaders can bring about behavioral change and the long-term adoption of new technologies.
The problem with many technological innovations is that they are solutions in search of problems. Organizations often bring in new technology without a clear idea of the use case for it, or they dream up new use cases to accommodate the innovations they aim to deploy.
This clearly is wrongheaded. Without a strong case for its relevance to the organizational mission and where it sits in the context of how the firm does business, a technology implementation is dead on arrival. Tech leaders should define how the new tool or platform will fit in the flow of the organization’s strategy and operations, down to workers’ daily routines.
Again, AI provides a useful lens through which to think through this challenge. At a typical insurance carrier, the efficient, proper and accurate completion of forms is a crucial aspect of operations. Even routine tasks can require numerous forms that must be done in a certain order with specific information. It’s no surprise, then, that users—employees, brokers, customers—often need help to fill out the right forms in the right way, at the right time.
This is an ideal use case for AI-powered solutions. Forms are a crucial part of operations, yet assisting users in their completion is often a somewhat mechanical process that nonetheless takes up the time and attention of a support staff member. A strong case can be made for an algorithm—one that’s capable of interpreting a user’s intent and guiding them through a series of prescribed steps—to take the burden of routine support tasks off workers’ plates, thus freeing them to devote themselves to more complex requests.
The lesson is unequivocal. For adoption to be successful, the innovation must integrate into existing use cases for which there is value to be unlocked in enhancing capabilities. It’ll be easier for the entire organization— technologically, structurally and operationally—to accept change, which in the best of cases can be challenging.
Adoption Can't Happen in a Vacuum
As I often tell clients implementing our AI digital assistant platform, the impact is revolutionary, but the implementation is incremental. Adoption does not happen overnight.
It also entails more than simply dropping a new technology platform or solution into an organization and turning it loose. Just as it’s a process, adoption of new tech can’t happen in a vacuum.
Organizations are made up of individual people, after all, each with their own preferences and proclivities. They also have institutional memories, and, for better or worse, specific approaches to operations to which they are accustomed. Leaders must keep both in mind to successfully implement new technology.
Dr. Sindhu Joseph is the founder and CEO of CogniCor, an AI-powered digital assistant platform for financial firms. She earned a Ph.D. in AI, holds six patents and is an author and speaker on topics around enterprise AI and the need for diversity across the tech industry.