Five years ago, the thought of adding virtual agents and the rigidness of their response seemed counterintuitive to delivering a great customer experience in the contact center. That is no longer the case, in fact, not deploying a virtual agent strategy is a hindrance to an exceptional customer experience.
The demand for self-service from your customers has grown to a level that cannot be ignored. Yes, there will always be customers who prefer the personal touch, and that should always be an option. That is what we are talking about most with our customers, options and letting the customer choose the path that best speaks to their definition of a great experience.
In an era where self-service and automation are in demand and define an efficient customer experience, the Intelligent Virtual Agent (IVA) is here as a real option to deliver that experience. While this technology has been evolving for the better part of a decade, the leaders of the contact center world have their versions, so which one is best?
At InflowCX, we know that there is not a one-size-fits-all solution for every company. So instead of asking which IVA is the best, we propose a few interview questions to ask the virtual agent and see if they pass to the next round of the interview!
1. Can You Understand and Process Natural Language?
The phrase “Natural Language” doesn’t really seem, well… “Natural”? Like learning a foreign language, we first learn the “proper” way to speak, which is not how humans speak. There are nuances, slang, and abbreviations to account for.
The real question for an IVA is to consider if it can pick up on those nuances. Think about how we speak in our day-to-day lives and the difficulty of processing that information without being in person. How many times has an email or text message been misunderstood? For any IVA, understanding the user’s conversational and even regional nuances is crucial. It means not just recognizing words but grasping the intent behind them.
The Intelligent Virtual Agent from Five9 does well in understanding natural conversation and what the user wants to accomplish. From basic tasks like password resets to more complex duties such as setting appointments, their IVA offers smooth conversations.
2. How Well Can You Multi-Task?
Real-life queries can be messy. Customers might have several questions that need addressing simultaneously, with different priorities. This is where an IVA can be very beneficial because we as humans get frustrated, and overwhelmed and may rush to the urgent versus the important task at hand. Removing emotion from the conversation has its benefits, and in multi-request scenarios, an IVA defers to action over emotion.
Logic and processes need to go hand in hand, and a virtual agent should be able to utilize logic to take care of the more important tasks first while not neglecting the other needs as well. Replicant’ does well in this area with its “Thinking Machine”, which interprets customer queries before deciding the next step. While that may feel like it could take a while, all the data and information are processed in fractions of seconds.
Analyzing the data that you currently have can help you prioritize how important this is, while also being able to quantify how many of these types of inquiries an IVA solution can help you resolve.
3. When Can You Start, and Are You a Fast Learner?
AKA: What’s Your Learning Curve and How Do You Adapt Over Time?
OK, that is two questions wrapped into one, however, it is important to consider how much training is needed for the virtual agent solution to get up and running. A new agent isn’t ready on day one, training and monitoring are involved. Taking that same approach with an IVA is needed to ensure success.
There is no off-the-shelf solution with virtual agents, and that is a good thing because just like a human agent, they need to absorb all the information about your company to become a resource for your customers. The onboarding process for a virtual agent and the seemingly endless capacity to retain information requires a different approach. The deployment of the IVA depends on the quality of historical data your company has.
IVAs should not remain static once deployed, this is not a “Set It And Forget It” solution. The resolution rate of an IVA is not 100%, however, it is something that should incrementally increase over time as it learns from the situations that were not resolved. Replicant, by harnessing the power of Large Language Models (LLMs) and the data from millions of interactions, has achieved an impressive 90% resolution rate. Their approach of constant refinement has led to a consistently improving virtual agent experience.
4. How Do You Integrate with Our Existing Systems and Data?
Like “how well do you work with others” the virtual agent solution needs to play nice with your current systems. Integration with the top CRMs like Salesforce and Hubspot is critical. It would be like having 90% of your workforce refuse to update the CRM, which just can’t happen.
A key component of a great customer experience is letting the customer know they are not just a number; you know them and they are important to your business. Genesys excels in personalizing interactions. By integrating existing data, their virtual agent solution can make the “virtual” feel incredibly “personal”.
With all the solutions you currently have in your contact center, now is also a good time to evaluate if there are redundancies that you could make more efficient. One of the many places to start would be your chatbot provider and ensuring your IVA integrates and plays well with that solution.
5. How Do You Handle Complex Issues or Escalations?
The current best-in-class virtual agent solutions are at a 90% resolution rate, so the question is not if they will need to ask for help, but how that transition will go. That transfer needs to be recognized as early as possible for the IVA and with little to no frustration. In other words, the last thing we want is the customer typing or shouting “live agent!” trying to get the appropriate help.
Industry leaders like Genesys, Zoom, Five9, Replicant, and NICE CXone, all emphasize a human-AI partnership. If a query is too intricate for the virtual agent, the customer is seamlessly transferred to a human, ensuring a smooth transition.
Part of that smooth transition, when a conversation needs human intervention after beginning with a virtual agent, is making sure the relevant information is transferred seamlessly. Regardless of being virtual or live, this approach guarantees that customers don’t have to repeat themselves, making multitasking efficient which is always the best practice that should be followed.
Virtual Agents Should Enhance Not Inhibit Customer Experience in Your Contact Center
A well-implemented, strategically aligned virtual agent solution can be a catalyst to scaling your business and contact center effectively. IVAs can help solve some of the largest challenges facing contact center operations such as staffing, consistency, and meeting customer’s omnichannel demands.
InflowCX has been rising to the challenges contact centers face for decades. We understand the contact center because we come from the contact center. Our approach always starts with your business first, contact us today to see why the best-in-class contact centers trust our team to deliver an exceptional customer experience in their contact center.
