The first rule of how to use automation in customer service is simple: don’t frustrate your customers.
Living up to this rule, however, is easier said than done. In fact, customer service has historically failed miserably, garnering a reputation for being incredibly frustrating and time consuming. This reputation has only become more ingrained as consumers get accustomed to seamless digital experiences and rapid, real-time communication. At the same time, those expectations of real-time, seamless communication can often prove challenging for brands to meet, much less exceed.
Now, however, brands finally have an opportunity to use automation at scale without frustrating customers. We’re not talking about cumbersome interactive voice response (IVR), or turning your CX over to an army of bots; we’re talking about using a combination of human empathy, bot-based automation, and AI-powered ticket classification to give every customer the best experience possible, without breaking the bank.
How to Use Automation Without Frustration
Today’s customer service automation can do a lot, but it can only go so far. At the end of the day, humans are still far more capable when it comes to handling ad-hoc or atypical requests. The key to acknowledging this in customer service is identifying the use cases that automation can handle faster and more effectively than humans, and those use cases that can benefit from a human touch.
Any customer service communication flow includes the following steps:
- Information collection
- Ticket classification based on category
- Ticket resolution
- Feedback collection
If we look at these four overarching steps, three out of the four (numbers 1, 2, and 4) can usually be handled through automation more effectively than they can be handled by humans. Information collection, for example, can easily be handled with a simple decision tree bot. For instance, if a customer is having issues with shipping, the bot can ask if the customer has the tracking number or not, and if so, what it is. If a customer is having an issue with the app crashing, the bot can ask if the customer has tried restarting the device, and if the app is updated to the latest version (this information might already be available through the SDK). Every category has essential but basic questions that a bot can automatically ask.
Similarly, ticket routing can easily be handled by an AI engine that automatically categorizes tickets. CSAT is almost always automated these days anyways, but by having a bot collect feedback within the messaging conversation, there is a higher likelihood of conversion. These three aforementioned steps are perfect use cases for frustration-free automation.
The third step, full issue resolution, is a bit more complex: some questions can be routed to a knowledge base article. Others can be solved through simple third party integrations (e.g. a tracking number for someone who can’t find the status of their package). Still others need a knowledgeable customer service agent to delve into the ticket and provide a solution.
That’s why this third step should not be relegated to only humans or only automation bots. Some issues require a human touch, and others do not. The best way of handling this is to offer “escape hatches” for customers to speak with an agent if the bot is not able to resolve an issue in its entirety.
These escape hatches are necessary — nobody wants to go back to the days of screaming at a phone to be connected with a human — but they are actually used less often than you’d expect. A recent Helpshift Benchmark report found that some brands were fully automating up to 70 percent of issues and using bots to assist agents by collecting upfront information in 20 percent of issues. This frees up agents to take care of the remaining 10 percent of highly complex issues, and despite the majority of tickets being automated to some extent, CSAT remained unaffected.
The reality is, customers don’t need to speak to an agent right off the bat in the vast majority of cases. And, as evidenced by unchanged CSAT after companies started to use automation intelligently, customers know that they don’t need to always talk with an agent; they just want their issues resolved in the most efficient manner possible.