When it comes to improving customer experience, we’re in the midst of a true revolution. Nearly 80 percent of companies selling to consumers report that they are already using or plan to use an AI-powered virtual agent for customer service by 2020. But implementing the technology is just the first step toward finding success.

Companies who want to make the most of a customer experience strategy that is augmented by artificial intelligence (AI) and includes chatbots need to identify and then track metrics that confirm their new technology is paying off. Doing so allows you to not only justify the investment, but also highlight trends and find areas where your virtual agent performance could improve. While the right metrics for your organization depend on your business and what you want to achieve, there are some overarching things that most organizations monitor.

Here are five key performance indicators (KPIs) to keep in mind as you work to automate and enhance your overall customer experience:

1. Customer engagement. The point of AI-augmented CX is to help customers more quickly find answers to their questions and address their needs. Tracking engagement KPIs can help you assess how customers are interacting with your chatbot—and the success of those interactions. Now as some have pointed out, you need to determine what successful engagement means to your company to craft a meaningful metric. For example, instead of the sheer number of chatbot interactions per conversation, measuring engagement could also entail tracking clicks on links to resources provided by the AI agent or monitoring whether customer issues are resolved without human involvement.

2. Satisfaction rates. Sometimes KPIs can be misleading. For example, longer interaction times with your virtual agent could mean that your customers are highly engaged – or that it’s taking too long for them to get what they need. This is where satisfaction rates come in. Give your customers the opportunity to rate their experience with your chatbot. The system can be as simple as a star rating or entail more involved questions. Many companies also offer a small incentive to customers, such as a gift card or discount, for providing feedback.

3. Confusion triggers. In an ideal world, your chatbot will provide the right response to your customers every time. However, the more likely scenario is that your chatbot may occasionally get confused by what your customers are asking or requesting. If the virtual agent can’t figure it out, then the inquiry will likely be routed to your customer service team. Tracking when and how often your chatbot responds with “I don’t know” or “I don’t understand” can help you improve the technology to reduce confusion. That, in turn, saves your staff time and creates for a better interaction overall.

4. Conversions. Your chatbot can do more than answer questions; it can help move customers toward an actual purchase. Facebook’s branded messenger bots provide a great example. The longstanding flower company 1-800 Flowers used a branded messenger bot on the social media platform to make it easier for customers to order flowers. The company’s CEO told Digiday that shortly after launch, 70 percent of their orders were from new customers, who connected with the bot and then bought flowers. Track your conversion rate to analyze the effectiveness of your virtual agents—and perhaps find new opportunities for more sales.

5. Cost efficiencies. Implementing a virtual agent not only makes your CX more efficient, it can also help your company better manage costs. Frost & Sullivan reports that chatbots typically “…deliver significant cost reductions, with ROI associated with some providers typically in less than twelve months.[i]” Consider the various cost savings opportunities with your virtual agent, from reducing staff time spent on redundant tasks to increasing customer contacts without hiring new employees.

Virtual agents are becoming part of the customer experience equation. Creating a plan for analyzing, measuring and improving on your chatbot experience can verify that your organization is making the most of this technology. 

Learn more about how AI can help you deliver personalized customer experiences.

Alla Reznik is Director, Customer Experience (CX) for Verizon's Global Products and Services group. She leads teams fulfilling the complete CX product life-cycle process and PNL, encompassing: concept and design; competitive analysis; branding and marketing strategy; and final development and deployment.

^[i] Stratecast paper by Jeff Cotrupe, Frost & Sullivan.