Venture capital investment in artificial intelligence increased by 72% in 2018 (source: PwC). And the term “artificial intelligence” was referenced 1,295 times in public companies’ earnings calls during 2018 (source: CB Insights). By comparison, uber-hyped “blockchain” was mentioned 1,028 times.
If those 2018 numbers are a leading indicator, 2019 is the year we expect a significant increase in the commercial adoption of AI. That includes the realm of customer support, as leading companies are implementing AI-powered virtual agents as a means of providing faster, better support that scales as their businesses grow.
Numbers aside, don’t panic if your company is just now vetting the idea of using AI and virtual agents for support automation. There’s some benefit to having waited this long. While you’ve been reading AI-related blogs in your favorite tech publications, the tech industry has learned a lot over the past few years. Even in its failure — as we wrote recently, we’ve identified “Four intelligence gaps that’ll kill your AI initiative.”
Meanwhile, the technology and key players that power AI have matured and the value of some use cases — including customer service and support — has become much more clear. So, how do you know whether AI-powered virtual agents are a good fit for your customer support network?
Our recent co-authored report with Contact Babel, “The Inner Circle Guide to AI, Chatbots & Machine Learning, offers a broad perspective on the state of technology and AI in customer support today. Among the topics we address is how to get started with AI.
If you’re considering implementing a virtual agent, these are some questions that you should ask yourself first:
How quickly do you need a virtual agent implemented, and what initial and ongoing resources will be required to ensure it’s successful?
There are talent, data, resource, launch and maintenance costs to consider. And it’s absolutely crucial that companies understand the ongoing work involved with launching and training an AI-powered virtual agent. As Ben Rigby wrote recently, AI needs human training to become truly intelligent.
Is there a specific pain point or issue within your operations that needs to be addressed?
Every technology project — whether AI is a part of the solution or not — should start with a business case. Define the problem you are trying to solve and the expected return on investment if you dedicate resources to solve it.
How will implementing an AI-powered virtual agent affect the overall customer experience, and how might the customer like this to be improved?
Your business is likely motivated to automate support to cut costs. But a well-executed virtual agent must provide big benefits for your customers too. For example, our customers have seen a substantial improvement in issue response time and resolution speed. Solving problems quickly is what matters most to your customers and virtual agents can play a big role in accomplishing this.
Are there AI solutions in the marketplace that have successfully addressed your specific issues in live environments?
Like any emerging technology segment, there are many different solutions and vendors that enterprises have piloted and invested in. Our recent report is a great place to start in understanding the different ways AI can be applied to customer support.
How would a virtual agent implementation disrupt existing operations?
AI-powered virtual agents are assuming an important role in today’s support landscape, but AI needs to be part of a cohesive multi-channel customer experience strategy. In particular, companies need to create a clear escalation path so that there’s a seamless handoff from AI to human support when necessary.
Are the improvements that AI promises measurable?
Yes, and your project should include clearly defined and measurable success metrics. In a recent post, we covered some of the ways companies measure the performance of support. Chief among them: self-service resolution, contact resolution rate, and the overall impact on contact center volume.
Is there a sufficient volume of data necessary to train an AI system effectively?
The more data you have, the smarter your virtual agent will be “out of the box.” But your AI also needs ongoing fresh content to understand new issues as they arise, along with a plan to provide continuous training so your AI receives the feedback signals it needs to become truly intelligent.
Will our infrastructure or existing platform need to be replaced?
It depends on the vendor, but the answer should be ‘no’. In addition to having a clear strategy for how your AI fits into your customer support processes, you need to understand how it’ll work with your existing technology stack. Make sure to invest in a virtual agent that can seamlessly integrate with your current infrastructure.
Is an AI-powered virtual agent the most effective way to address your needs?
You should never invest in AI just for the sake of saying you did (or because your CFO mentioned it on an earnings call). Make sure that the virtual agent has the potential to solve the specific use case(s) you are trying to address.
Have any trouble answering these questions? Let’s talk
It might be difficult to answer some of these questions without a deep understanding of the machine learning technology that powers AI and what a virtual agent might look like in practice. That’s where we can help. We work with a range of technology partners and help our customers solve their support challenges with the right AI-powered solutions. We’re happy to talk with you about your specific needs, give you a demo of an AI-powered virtual agent in action, and offer up advice — even if our CX automation platform isn’t the right fit for you. Contact us to set up a free consultation and demo.
Download our report today
Make sure to download our full report: “The Inner Circle Guide to AI, Chatbots & Machine Learning.”