What is a Virtual Agent?

A virtual agent, also known as an IVA, virtual rep, or chatbot, is a software application that uses natural language processing and pre-defined answers to support humans online.

This article outlines a number of aspects of virtual agents, with the goal of making it easy for you to adopt the technology if you believe it will serve your organization.

Virtual Agents Explained

Intelligent virtual agents are computer programs that leverage a mix of programmed rules and conversational artificial intelligence in order to offer a simple service or provide basic help. Virtual agents are a larger category of online services, containing chatbots, voice bots, and even interactive voice response systems.

In many cases, it can be helpful to think of a virtual agent as a digital assistant, particularly if you're thinking about the agent as a way for your business to serve users. To this end, virtual agents are most commonly used by companies who serve a large customer base or who have a large number of internal employees.

Although virtual agents, chatbots, and AI are often used interchangeably, it is important to note that a virtual assistant and a chatbot are similar but not the same. The term chatbot refers to a specific type of virtual assistant designed to communicate via chat (via email or messaging). Virtual agents, on the other hand, can communicate through any given medium — for example, via voice response over the phone.

The last note here: be careful not to confuse virtual agents with virtual assistants. A virtual assistant (VA) is a remote customer service role that uses technology to provide service without the customer's need to speak to the agent. It's similar to a virtual agent because the agent only interacts with the customer via technology.

Virtual Agent Benefits

Virtual agents can provide a number of benefits to the organizations that use them. Specifically, there are three main reasons why a company might use a virtual agent:

Consistent, High-Quality Answers

You can count on your virtual agent's answers to be correct and on-brand 100% of the time. In this way, virtual agents can help you to delight your customers and increase your organizational efficiency.

Improving Agent Efficiency

Virtual agents free up your human agents' time by handling simple requests, which have predictable answers and the highest total volume. The key here is for virtual agents to complement your human agents, not replace them. 

At Ada, we've built tools that make the hand-off from ai to human quick and painless so that customers with complex inquiries get the support they need.

Reducing Response Time

People hate spending time on hold. One of the easiest ways you can delight your customers is by responding quickly, which is exactly what your virtual agent is best at. 

How Virtual Agents Work

While implementation varies, most virtual agents require that organizations map out specific workflows that a virtual agent should handle.

When a customer request is made, the virtual agent uses natural language processing to identify the keywords that a human customer used in that inquiry. The virtual agent then interacts with its scripted responses to offer the best response to that inquiry.

The most advanced virtual agents move beyond simple keyword identification and use conversational AI to extract and carry context throughout a conversation. Of course, many lower-end chatbots will fail to adequately understand and remember this context.

This conversational AI is really what sets a good virtual agent apart from a bad one. The team at Ada has invested thousands of hours in ensuring your bot seamlessly offers up answers from your customer support knowledge base. Proprietary improvements in conversational AI allow our bot to delight customers with a quick resolution.

For example, if a customer makes a comment about liking a certain brand of clothes, robust conversational AI might recommend similar affordable alternatives for the customer to consider. Similarly, if a customer complains about a bug, high-quality conversational AI might recognize that the user's settings have caused this error and explain how to address the situation.

To ensure you're getting the best performance, it's imperative that your virtual agent has a robust system in place for measuring and monitoring response quality. By incorporating user feedback, your agent responses should be able to continuously improve over time.

Virtual Agent Use Cases

Virtual Agents for Customer Service

The most popular use case for a virtual agent is to help with customer support. Specifically, virtual agents can be robust tools for answering routine customer questions, triaging common complaints, or actioning regular requests.

And don't just take our word for it; study after study has shown that virtual agents add value to customer service functions. At Ada, we've seen that companies consistently reduce the hours spent on customer requests by over 80%. In turn, their support teams have far more time for more complex tasks.

Here's an example to demonstrate the above. If you operated a telecommunications company, you might have all of the live chat inquiries funnel through a virtual agent before redirecting complex requests to your trained customer success team.

Virtual Agents for Marketing Automation

The number-two most popular virtual agent use case is that of lead qualification and triage. You've probably encountered at least one virtual agent that offers to collect your details & pass you on to the right team member online.

Not only do virtual agents of this kind help customers quickly achieve their goals, but they can also help businesses to parse through low-quality inbound requests. Everyone wins when your sales organization avoids a high volume of low converting leads.

Virtual Agents for Employee Support

More recently, virtual agents are starting to see significant use in helping with internal organizational processes. For example, a large enterprise's IT team might use a virtual agent to support employees who need help resetting their passwords, setting up new computers, or accessing shared accounts.

A related application is that of employee onboarding. If a new hire needs to complete a predefined onboarding checklist, a virtual agent can quickly guide them through their to-do's without requiring that they ask a peer for a favor.

Similarly, if an employee recently transitioned to a new department, a virtual agent might help them quickly understand the tools & processes involved with their team. Here, the virtual agent is playing the role of "mentor" or "buddy," — helping with knowledge transfer and inculcation.

It's important to note that any given virtual agent's ability is bounded by the data and models powering their natural language processing. Getting these things right is so hard that we've hired dozens of the best minds in deep learning from across North America at Ada. 

Setting up a virtual agent

While a virtual agent does not always need to be trained, companies need to consider the ramifications of virtual agent adoption. Older companies that use on-premises deployments can require months of time to train and launch your agent.

We've heard firsthand how frustrating this experience can be, so we designed our virtual agent to be deployed with a single line of code that you can add to any page on your website.

Training a virtual agent

Similarly, other companies often leverage a complex process to their virtual agents. They might ask you to dig through customer support requests, create rules to process the requests, and populate the rules with text samples

At Ada, we take an alternative approach to training. We base our bots on your existing customer support knowledge base. This way, you don't need to wrestle with any additional inputs to get started.

Testing a virtual agent

Once your virtual agent is built, we highly encourage testing on a small group of customers to ensure that it operates as expected. Once you've evaluated and corrected any bugs, you can proceed to deploy the virtual agent more broadly. Customers will provide a wealth of data in the form of feedback, allowing you to keep your virtual agent up-to-date and relevant.

Responding to inquiries across channels

In many cases, virtual agents have now begun to incorporate the ability to resolve customer, employee, or corporate work requests across multiple channels, including voice, text, and chat, as well as through digital assistants, multi-channel applications, and email.

The best conversational ai platforms can connect with back-end systems so that when a user requests a piece of information or help, the chatbot can also connect back to your ERP or CRM, for example.

At Ada, we've spent a significant amount of time investing in these third-party integrations so that you can centralize your customer support requests & triage responses from the channels that matter most to you.

The virtual agent market

Better yet, Gartner predicts that by 2021, 25% of digital workers will use a virtual assistant daily in their tasks, while by 2020, 25% of customer service and support operations will use a virtual assistant across their engagement channels, up from 2% in 2017. 

With COVID, this past year has seen an explosion in virtual agents' popularity, reaching number 1 in several industry lists citing fast-growing technologies.

The field is expected to see continual growth, with one study estimating that the global market for virtual assistant software will reach $17 billion in the next two years. Since 2016 we've seen a flurry of IVA offerings grow in prominence.

Amazon's Alexa, Apple's Siri, Microsoft's Cortana, and Google's Assistant, as well as Facebook's M and Samsung's Bixby, are just a few of the programs that have cracked the top 10 list of most popular virtual agents.

We've also seen virtual agents from smaller companies emerge, including Vika and Amy, two virtual agents that can have real-time conversations with users. In general, these newer entrants tend to focus on consumer use cases, as our team at Ada and other established companies have well-addressed the need for enterprise-scale virtual agents.

You might be wondering why the sudden interest in virtual agents. Well, the ubiquitous rise of voice-controlled assistants, such as Siri and Alexa, has popularized the idea of using digital assistants to help with day-to-day tasks. 

More recently, virtual communication surged over the course of 2020, as remote work was thrust on much of the world. During this shift, companies needed to increase their digital employee satisfaction and reign in a dramatic increase in online customer service requests. Virtual agents have allowed them to do both.

How can a virtual agent drive revenue for my business?

If you're a business with inbound customer requests, a virtual agent can help you to increase your MQL to SQL conversion rate, drive more appointments for sales, and drive up visit to lead conversion rates by allowing customers to convert on every page of your website, rather than just the sign-up page.

Perhaps most exciting, a well-run virtual agent can actually improve your lead-to-customer conversion rate. This typically happens in one of two ways. First, your virtual agent might allow you to collect more personalized information from your prospects. In turn, your sales agents will have more context and can be more persuasive on the phone. Second, virtual agents can often reduce your time to call by helping you screen-out low-quality prospects, giving your sales team more "uptime" to quickly call-back.


How can a Virtual Agent Reduce My Costs?

As described above, at Ada, we've consistently seen that customer support automation can double-digit percentages of all incoming customer inquiries. In doing so, we often see organizations increase the overall quality of their customer support, as the added time allows human agents to be more thoughtful in their responses to complex customer inquiries.

What are the top-rated virtual agents?

The leading customer support virtual agent is Ada. From Zoom to Facebook, we've helped hundreds of companies automate customer support. All of this data helps ensure that we can build delightful virtual agents in industries ranging from travel and hospitality to technology and banking.

The leading virtual agent for recruiting is either paradox.ai or HireValue. Both have a large focus on candidate experience and are purpose-built for human resources teams at larger organizations.

The top Azure Virtual Agent is, unsurprisingly, Microsoft's Power Virtual Agent. They've built out a comprehensive suite and rolled it out to organizations including Kobe and MyGov.

The top inbound sales virtual agent is also Ada. Rather than viewing the customer journey as a fragmented funnel, Ada takes a holistic approach to customer experience that allows you to better convert customers at every step of the funnel.

How much do virtual agents cost?

Lower-end virtual agents might quote you with an off-the-shelf price, but at Ada, we want to ensure we're delivering a solution that's tailored to your organization.

To get a personalized quote based on your use case and volume, talk to our team.