Spectrum of Care
August 13, 2020 • 4 minute read

Live Agents and Virtual Agents: The Spectrum of Care

Imagine yourself on a busy day trying to cancel an appointment and you keep getting put on hold waiting for a live agent. You know your task can be completed through self-service–it’s just a simple cancellation. After 30 minutes of waiting you finally get put through to an agent, who is able to quickly cancel your appointment. You are frustrated because you had to spend 30 minutes on a task that could have taken two due to the wait time.

Or imagine a scenario where you are trying to get a refund for a damaged item past the allotted refund date. You want to speak to a live agent to explain your situation and bypass business rules. However, there is only a self-service system in place to help you. Your attempt ends with having to go to the actual store to get the refund. Again, your time and effort are wasted.

Chances are, you’ve experienced a transaction like this that was not properly handled. Either there is too much automation or not enough. 

As virtual assistants become more advanced, they are able to take over more and more transactions and offer positive customer experiences on par with live agents. Tasks such as resetting a password can be effortlessly completed with automation without compromising the level of service. However, there will always be a place for human agents, and the key is knowing when to use which (or both). When a conversation flow is properly designed, a company will be able to handle each transaction type in a way that is cost-effective and results in the highest possible customer experience.

Keep in mind that some customer conversations may include more than one transaction, and therefore more than one transaction type. For example, a customer may call their healthcare provider in order to schedule another appointment, reset their password, and pay their bill. 

Types of transactions

We can categorize transactions into three main types: 

  • Simple transactions
  • Repetitive, data-driven transactions
  • Out of the ordinary transactions

Simple transactions are best suited for automated systems. Examples include tasks such as inquiring about operating hours, directing a call to a certain department, checking an account balance, or other FAQs. These transactions can be handled by a simple chatbot or IVR system, as not much advanced functionality is needed. However, the level of customer experience can decline significantly if the system is too limited in functionality, so using a more advanced Conversational AI system will result in higher customer experience. With an automated system, a high-volume of these transactions can be handled at a low operating cost. These simple transactions 

Repetitive, data-driven transactions are the most common transaction type. They are best handled by an advanced automated system, such as an Intelligent Virtual Assistant (IVA). Examples of this type of transaction include updating account information, identifying and authenticating customers, paying bills, enrollments and onboarding, and booking reservations. These transactions are more conversational than the simple category, but can still be properly handled with automated systems. Due to their more complex nature, it’s important to have an automated system that is advanced enough in order to maintain a high-level of customer satisfaction. With the proper system in place, operating costs can stay low even when the volume increases, like during seasonality spikes and crises.

Last but not least, out of the ordinary transactions are those that are best suited for a live agent to be properly handled. As mentioned in the opening example, this transaction type usually involves a bypass of business rules or an extenuating circumstance. These transaction types can not be easily scaled due to their individualistic nature, and therefore tend to have higher costs per transaction. However, a live agent is needed in order to result in the highest level of customer experience for this transaction type. 

Choosing the right level of automation for your business

A deep knowledge of the customer experience space and conversational design is needed to determine when a live agent, a virtual agent, or both should be used in a given situation. And depending on the company, the industry, and your customer base, the same transaction may be better suited for different levels of automation. 

Especially because the most common transaction types are repetitive, data-driven transactions, it’s important that the automated system that you deploy be powered with advanced Conversational AI technology. Working with a partner who understands the importance of both human and virtual agents in customer experience settings is essential to get the balance right. 


Learn more about the importance of conversational expertise and working with a managed service in our whitepaper.


Want to learn more? Let’s talk.