Increasingly, enterprise service desks are adding artificial intelligence and automation to their operations. It’s a decision that can speed up ticket resolution and increase CSAT scores.
At Alphanumeric, we work closely with clients who want to integrate AI into their service desks. One question we hear a lot is this:
Do customer service KPIs change with AI and service desk integration?
It’s always a good idea to revisit your customer service KPIs. Let’s talk about some of the ways your KPIs might change after AI integration.
Traditional Customer Service KPIs
When measuring the performance of your service desk, there are many key performance indicators you can track. Some are universal while others may be specific to your industry or even to a product or service you sell.
If an enterprise tracks only a few KPIs, they are most likely to be:
- Occupancy/Utilization, which measures the percentage of a customer service representative’s time that is spent doing things like answering tickets, chatting with customers, or taking phone calls. It typically does not include training time. A typical goal for CSRs is 75% to 85% utilization.
- Productivity, which measures things like the number of tickets handled per hour or the number of tickets solved per hour. It’s a good measurement of how your CSRs are managing their time and how efficiently they’re working.
- Customer satisfaction, which measures how happy your clients are with the service they receive from your CSRs.
Your service desk will still have human agents after AI integration. Since that’s the case, you’ll need to continue tracking your core KPIs to measure their performance and ensure they’re working efficiently.
KPIs for Artificial Intelligence
As you might expect, the KPIs for artificial intelligence are not the same as those for human agents. One of the biggest differences is that chatbots and virtual agents don’t need to take time away from “work” for training. Their training happens in the form of machine learning and script writing.
The key, then, is to track metrics that allow you to measure how well your AI solutions are working for your clients. That requires some lateral thinking and creativity. Here are some KPIs to consider.
- Customer engagement. You will probably want to measure how many users are engaging with chatbots or virtual agents. Beyond that, it’s a good idea to track specifics such as how many users click on AI-provided links or how many tickets are resolved without human interaction.
- Customer satisfaction rates. One trap that enterprises can fall into with AI is assuming that a long customer interaction equals satisfaction. Instead of guessing or assuming, give clients who interact with AI the opportunity to provide a rating at the end of their interaction.
- Conversions. This metric only applies if your AI solutions are designed, at least in part, to drive sales. If they are, you should measure how many of the people who use chatbots or virtual agents go on to make a purchase. It’s a good way to assess the profitability of your AI.
- Confusion triggers. Any time an enterprise introduces AI to a service desk, there’s a learning curve. NLP and other tools can minimize confusion but it’s still likely that a chatbot will occasionally say “I don’t understand” or “I don’t know.” Those things can be frustrating and stressful for customers. Tracking them gives you the opportunity to make programming and scripting adjustments that will improve customer satisfaction.
- ROI. One big reason that AI integration is on the rise is that in the long run it can save enterprises money. It may be helpful to track improvements in efficiency and other metrics to get an idea of what kind of return you’re getting on your investment in AI.
These aren’t the only KPIs you can track but they are five of the most important.Integrating AI into your existing service desk doesn’t mean that you can stop tracking your usual KPIs. However, it does require the addition of KPIs that will help you evaluate and improve the performance of your AI solutions.To learn how Alphanumeric can help you integrate AI into your service desk, please click here.