Finnish telecommunications company Elisa recently replaced its functional chatbot — which answered customers’ routine questions — with an intelligent virtual agent (IVA) called Annika. Executives at the organization expect this sophisticated new robot to engage callers with more empathy and depth than its predecessor.
As Mailiis Ploomann, Head of Telecom Services at Elisa, told its partner MindTitan, “The most important thing for any telecom company is to be there for your customer exactly when they need a solution, whatever question or problem they may have.”
The company also hopes this robot will free time for Elisa’s human agents, helping them handle more complex customer inquiries and navigate higher-value activities.
This change comes on the heels of a rocky period for Elisa.
In 2017, it merged with Estonian cable television and internet provider Starman, forcing it to rethink its customer service.
Two years later, COVID-19 caused a spike in customer inquiries — alongside considerable disruption to its physical workforce — leading to 500 customer agents grappling with more than 100,000 incoming calls per month. On top of that, the number of topics Elisa customer reps needed to master grew from 400 in 2015 to 800 by 2021.
These issues, coupled with high natural staff turnover and the need to quickly train new agents, meant Elisa’s customer service department critically needed IVA-based support.
Why Intelligent Virtual Agents?
IVAs are sophisticated conversational robots, and they significantly contrast with the basic chatbots known for answering FAQs.
Chatbots follow a discrete number of rules for directing consumers to answers to the questions they’re most likely to ask.
For example, you may want to know how much a certain movie costs. The platform’s bot directs you to the relevant response. If you want to know more, such as how to get a price reduction, you’d have to reach a human agent.
Functional chatbots are severely limited, capable of handling only the most basic requests, and therefore satisfy far fewer callers than IVAs.
Related Article: 8 Examples of Artificial Intelligence in the Workplace
Intelligent Virtual Agents: Empathetic and Smart
IVAs, which are also called Intelligent Personal Assistants (IPAs), function on artificial intelligence (AI), machine learning (ML) and natural language processing (NLP). That means they’re programmed to simulate our brains to recognize, perceive, predict and judge.
The best of them — think Siri, Cortana and Google Assistant — ask questions to get to know you better. The most sophisticated IVAs know over 120 languages and dialects and can master contextual analysis.
IVAs cross industries, from IT to HR, offering multi-channel capabilities that leverage chat and SMS. In FinTech, for example, one of the most respected IVAs is Eva of HDFC Bank, which feeds investment recommendations, relevant real-time news and other industry-related data to customers.
The Benefits of Intelligent Virtual Agents
IVAs help you in ways that include the following:
- Improved customer service: Intelligent virtual agents field customer calls 24/7. That feature was particularly useful during the COVID-19 shutdown, when business usage of IVAs grew by 75%, according to IBM.
- Increased customer satisfaction: A 2022 survey from Tidio revealed that 62% of customers prefer speaking to an IVA over a human. There’s also reduced wait time due to increased efficiency, allowing brands to help customers faster and, therefore, retain them.
- Reduced costs: According to Cameron James, regional manager of Unity Connected Solutions, IVAs typically cost 20% less than a live agent. Plus, using these bots allows businesses to reduce customer service costs by up to 30% (Chatbots Life).
- Ensured compliance: IVAs are law-abiding robots that automatically adhere to HIPAA, PCI DSS and other industry standards and regulations.
- Increased revenue: IVAs answer repetitive questions and source information via backend integration, allowing agents to focus on selling and customer retention. Plus, they help agents understand customers better, increasing upselling and cross-selling and speeding up the purchase decision process.
On the other hand, “intelligent” doesn't mean intelligent like humans. These IVAs — incredibly smart though they are — can’t think, reason or make decisions since they ultimately reduce to distinct functionalities. Therefore, they’re limited in their problem-solving skills, lack moralistic reasoning and lag in understanding contextual information.
While tech leaders work to develop cognitive sophistication in these bots, the best available to date are market players (like customer experience platform BirdEye) that experiment with sentiment analysis, where intelligent bots analyze customer sentiment and emotion to determine how to respond.
Is an IVA the Right Fit for You?
Here’s how Ladd Wimmer, Enterprise Solution Consultant at Five9, answers that question:
If you experience any of the following drivers, you’re almost certainly a perfect fit for instituting an IVA. They are:
- You’re an inbound or outbound customer interaction center with a medium-to-large volume of calls. IVAs can enhance the inbound/outbound experience with self-service portal applications, alerts, appointment reminders and more.
- Your automated phone system frustrates new customers with its multiple options. Or, a growing number of customers request text-to-speech (TTS) technology or multilingual support.
- You encounter low IVR (Interactive Voice Response) containment rates, where callers leave in frustration after repeatedly pressing zero for an agent.
- You periodically navigate predictable spikes in demand, such as over holidays or high travel periods. Or, you experience an unpredictable spike in demand, such as with a natural disaster or product recalls.
When introducing the technology, Wimmer suggested you add an IVA to your existing contact logistics solution; there’s no need to forklift your current contact center solution.
Look for repetitive tasks that can easily be automated (e.g., routine FAQs, claims and orders statuses or address changes). You may also want to consider your company’s greatest pain points and associated quick fixes and automate those before expanding to add-ons.
Outside of customer assistance, other types of IVAs range from skill bots designed to obey human commands (like Alexa) to transactional bots (like Lyft autonomous cars) or informational bots (like Poncho the Weather Cat). The most sophisticated of these IVAs are conversational bots that mimic written or spoken human speech to an incredible degree.
Speaking of which, who can forget the famous Canadian hitchhiking robot called hitchBOT, a child-sized robot built from pieces like a beer cooler bucket, noodles and rubber boots and gloves, using Wikipedia as its source of information to tour the world — though it unfortunately met its end in Philadelphia.
Related Article: 5 Best Practices for Using AI in Your CX Strategy
The Bottom Line
What does implementing an IVA really look like?
For Finish telecommunications company Elisa, the technology handled 45% of all of its inbound calls — and was able to fully resolve 20% of them.
On top of that, Annika was more knowledgeable than any of its human agents since it mastered and could converse on more than 70 different topics.
That’s the beauty of IVAs.