Let’s be honest here—advanced AI technology such as digital labor carries incredible potential for increasing and sustaining an organization’s competitive advantage, but it’s not a “plug and play” technology either. The IT business environments where virtual agents can thrive are complex and as such, there are some adjustments that need to take place before “reaping the spoils”, so to speak. Similar to the way 15th century European explorers set out into a vastly unknown world seeking better trade routes, early adopters of virtual agents are also embarking on a journey into the unknown to gain greater insights into the application of AI within their business models.

Trailblazers like SEB Bank, UBS and other companies are forging a new path in the way business operations can be handled and at the same time, are recording their new learnings as they go. They’ve engaged in a pilot program that is in production to learn about the installation process of virtual agents and the different business areas they can impact once they’re implemented.

The early adopters I’ve mentioned above have shared some of their CVA (cognitive virtual agent) learnings and here are some of the more important ones to point out as other businesses consider initiating a CVA pilot program of their own.

Set your virtual agent up for success.

Across each of these use cases, clients stressed the importance of implementing your CVA where its cognitive capabilities can flourish and eliminate real pain points for employees, customers or both. Ideally, the pilot program should be centered around a service that will ultimately require some degree of enterprise scalability because its business impact will be more profound once a full CVA rollout is in effect.

Establish the collaborative dynamics of your hybrid workforce.

All the adopters used this exploratory program as a good time to determine how human and virtual agents would collaborate in order to enhance the customer’s overall service experience. Depending on the industry and business process, a virtual agent could handle a repetitive customer issue and escalate it to a representative when appropriate. In addition, adopters can also deploy virtual agents in strictly an assisted intelligence capacity to help human agents quickly pull up client information. These are decisions that should be made during this pilot phase so large-scale deployments can happen without complications.

Get the C-Suite onboard.

SEB Bank, a major financial institution, has so far observed that the strong support of C-level executives is vital in order to keep CVA projects progressing in the right direction. To garner this support, CVA project leads need to communicate to C-level executives that virtual agents are intended to be long-term business investments and generally do not provide short-term gains. The purpose of cognitive virtual agents is to transform the manner in which business operations are conducted so that companies can run more efficiently and cost-effectively. For such a technological shift to occur, executive-level management need to stand behind this transformative vision and amend their thought leadership in order to win over resisting parties.

Know your target audience.

An online gaming company quickly realized that given their particular customer service needs, the virtual agent they deployed would have to interact naturally with customers before it could begin to yield positive business results. After introducing a virtual agent to a select portion of their customers, mainly comprised of online gamers, they noticed a drop in customer satisfaction levels. For instance, when a gamer used words like “fuzzy” or “messy” to describe their issue, the virtual agent would ask them for his/her email address. By identifying these interactive, natural language discrepancies in this exploratory phase, and correcting them afterwards, customer acceptance rates of the virtual agent increased.

They employed an agile training method for their CVA—a series of trial and error actions—and gradually boosted customer satisfaction levels. Understanding the training demands of your virtual agent ahead of time will start to shape a company’s best practices for future CVA implementations.

Communicate the benefits of virtual agents to employees clearly.

SEB Bank found that informing employees about the benefits of a CVA in a way that appeals to their professional development was particularly effective in lowering their staff’s initial apprehension. Their messaging included the following benefits:

  • They would no longer be tasked with manual, repetitive work; virtual agents would now take this over.
  • Going forward, employees would handle higher priority issues and in the process develop new business skills.
  • The inclusion of a CVA into the workforce would naturally open up AI-centric roles dedicated to their ongoing maintenance and training.

Testing out virtual agents in a live yet controlled environment delivers first-time adopters a wealth of knowledge—from figuring out the best approach in introducing a digital worker to staff and customers to taking note of the obvious pitfalls in the implementation of a CVA. Ultimately organizations that are exploring the business value of AI and rethinking their approach to business operations now are maximizing their competitive advantage down the road. With the increasing influence and popularity of AI in business circles, the cost of not “testing the waters” with cognitive virtual agents might exceed the cost of investing early in digital labor.