The most effective customer service operations are built on efficiency. When there are peaks in customer service requests, it’s important that customers are dealt with as quickly as possible to keep things manageable. Ticket routing is at the heart of this type of organisation.
It’s especially important in financial services and governments where there are complex products being handled by different departments. But in retail and ecommerce, ticket routing is still important for any team looking to scale. And with the advances in AI, it can be made easier.
What is Ticket Routing?
Ticket Routing is the practice of allocating customer service requests to the most appropriate agent or department. What counts as “appropriate” depends on your business needs.
For example: you may have different teams managing pre-purchase questions and post-purchase questions.
Alternatively you may have particular agents who are better at different things or who prefer to handle different types of questions. That way you can be more sophisticated with it.
You can go a step further and rank tickets by priority and bump high priority tickets to the top of a queue. Or if a VIP customer comes through, they can have a dedicated agent or process.
Ticket routing is also important from the customer’s perspective. As a customer, you don’t want to be speaking to someone who has no idea how to help you and so passes you along.
How does it work?
If you’ve ever called a customer service phone number you’ll be familiar with the “Automated Assistant” which asks you to press 1 for sales, or press 2 for customer service and so on. That is a form of ticket routing, but just one where the customer does it themselves.
In the modern age, this has started to be replaced by Interactive Voice Response (IVR) where a customer gives a short summary of the problem, and the voice recognition software then assigns you to the most appropriate route. That’s one example where AI can come in.
With emails or contact forms, you can either ask customers to self-route using dropdown options or different email addresses. Or you can have someone in your team manually tagging different tickets in order to route them to the right place.
How does AI make Ticket Routing better?
If one of the goals in customer service is to help customers as quickly as possible, then having a manual routing process adds a bottleneck and potential delay that may slow down the process. When the person manually sorting looks at a pile of unsorted tickets in the morning, the ones at the bottom could be critically important, but going through one-by-one will mean a delay.
This is the first place where AI can help. Using natural language processing, a conversational AI engine can parse the ticket and tag it based on a number of factors: intent, category, and even sentiment, i.e. how happy or unhappy the customer seems based on the language used.
This can be done in seconds when the ticket first arrives, and then placed in the queue. Using AI can also help eliminate human error that arises when someone is just skim reading lots of tickets. The AI will read the whole message, and can be trained to follow procedures to the letter.
Process analysis
You could also use AI to assess how effective your current routing is. By looking at CSAT, first response time and any other metrics you care about for different routes, you can then find areas for improvement within your flows.
This can be useful for balancing things within your team. You may find that one of your team is excellent at providing product recommendations and gets great feedback, and so you’ll want to route as many of those queries to them. Or alternatively, someone is extra efficient at dealing with returns quickly.
Bringing AI into the ticket routing
The main way that AI plays a role is by actually responding to and in many cases resolving queries without passing to an agent at all.
Information gathering
In many cases when a customer asks something, it’s necessary for the agent to find out a little more information in order to answer the query. This might be an order number, or even pictures of a damaged item when it comes to a refund claim.
If the exchange happens via email, this back and forth of information can happen over days or even weeks. By bringing AI into it, the AI agent can assess whether all the information is there, and if not go back immediately to try and get what’s needed. Once everything is in place, the human agent can step in.
Resolving tickets there and then
For most retailers there are between 40-80% of tickets that could be answered without humans needing to get involved at all. Whether it’s questions about product features or delivery costs, all the way through to questions about specific orders and returns. If the AI is set up correctly and has the appropriate integrations with knowledge bases and your tech stack, then anything is possible.
Take an order that’s gone missing. If a customer asks about it, through deep integrations with the warehouse, order management system and carriers, it can locate where the parcel is supposed to be. Then it can ask the customer if they’ve received the package, and start the process of ordering a replacement or a refund depending on the situation.
That’s doing everything that an agent would do, in essence, but it can be done much faster.
How does this impact routing? The questions that are best dealt with by AI can be routed that way, while the ones best handled by a human can go that way.
For example, many retailers we work with automatically route any conversation about sensitive topics to a human. These could include if a return window has been missed because a family member has died.
Summarising tickets
Even if your ticket is resolved in the end by a human, AI can still have an impact on the way. Both through information gathering (described above) but also by summarising the conversation that the AI has had before the agent picks up the ticket.
Summarising tickets allows agents to get all the context rather than having to scroll through a long series of messages to work out where the customer is in the conversation.
These are just some of the ways that AI can help you to make customer service more effective. If you want more information about how you can use AI to make your customer service more effective, speak to our team today.
