AI-Assisted Agents Can Help You Cut Average Handle Time
Find out how to speed up your team's interactions without sacrificing quality.
November 20th, 2017
We all know that cutting down average handle time (AHT) is important. Customers want solutions fast and they want to spend as little time as possible having to explain their situation. A shorter AHT means your operators are working more efficiently. And that saves you money. But now chatbots can help you shorten your AHT even more. So here are our shortening average handle time tips:
Greetings & Data Gathering
It isn’t uncommon for businesses to train their employees on different greeting tactics. Leading questions and short and sweet hello’s are a couple of the many ways to shave seconds off an operator’s AHT. In the past, operators had to know how to best get a customer to give up information without any drawn-out small talk.
Now, chatbots can do all that work for the operator. By their very nature, AI-assisted agents don’t make small talk. You can pre-determine a range of greetings and give them to the bot. Then it can instantly send this message out whenever a customer contacts you.
Once greetings are out of the way, every customer service (CS) interaction requires some gathering of information. Why is the customer contacting the company? Do they have an order or account number? What is their name? This is all information that's useful – and often vital – to the operator if they’re going to solve the problem.
Traditionally, this information would be collected manually. An operator needs to ask for a full name or an account number. Perhaps they then need to search for the person in their database. Or maybe the customer doesn’t have an account so the operator needs to ask for an address. All of this information gathering takes time. Not so with chatbots.
You can program the AI to ask for all the appropriate information needed for that interaction. To do that, it has to recognise the type of query, what information it already has, and what information it still needs.
With complex technology like this, it's no wonder that chatbots are on the rise. By the time the agent enters the interaction, they have all the information they need to begin working on a solution. This means they can start with a simple ‘hello’ and move right on to finding a solution, if they haven't already.
And, because a bot is software and not human, it only needs ‘training’ once. It receives a program and then will continue on implementing that unless told otherwise. Humans, on the other hand, need regular refresher training to keep those skills sharp.
Knowledge bases are essential for operators to do their job properly. Operators can use them to look up any information that they’re not sure about. So, to that end, knowledge bases are still relevant today. But they can also be given to an AI.
AI’s are able to absorb and – more importantly – retain vast quantities of information. This means that once you provide an AI with access to a knowledge base, it simply knows that information. Searching for something becomes a simple task of analysing data that the AI already has – a matter of seconds.
And you can now create knowledge bases specifically for your AI. Some AI can draw data from previous conversations to suggest the best answer. Not only is the AI able to understand what kind of interaction it's dealing with, but it can also understand what the topic is and how much of a priority it is.
The AI needs a bank of information, full of historical data of your best interactions. It uses this to suggest a fully formatted answer to any initial query. The operator can simply press 'send' or choose to edit the answer. Either way, this technology can shave minutes off your AHT.
But wait – there's more! Chatbots will integrate any new additions or replacements into the data they already have. As soon as you update any processes, the chatbot will automatically implement the changes. Operators, however, will not. With an operator, this kind of thing needs to be communicated on a wide scale.
Emails, notifications, social posts, they can all be ignored or missed. FM implements weekly quizzes just to make sure our operators are always on top of the latest business changes. So, although this factor doesn't necessarily reduce your AHT directly, it can optimise the overall efficiency of your team.
It’s important that the same agent deal with an interaction from beginning to end, as long as the interaction doesn't have to be escalated for any reason.
Agent consistency allows an operator and a customer to build a good rapport. If the customer likes and trusts the operator, they’re going to be much more open to advice or solutions. And, generally, they’re going to be having a better customer experience (CX).
But it isn’t just about being friendly.
Switching operators means the new agent has to delve into all the appropriate data for this interaction. Although AI can now make that data easy to access, it still adds time to any conversation. And a new operator could very easily end up asking the same questions twice, needlessly increasing their AHT and frustrating the customer at the same time.
So where does automation come in? AI-assistants can be programmed to assign the same operator to any follow-up interactions. Not only that, they can also automatically send follow-up emails, status updates, surveys, and so on. And if the customer replies to any of these? The AI can once again assign that same operator to respond, along with all the information they need to keep their AHT as low as possible.
Call routing is similar, in some regards, to agent consistency. It’s all about getting that customer to the right place. But it's also more about sending the customer to the right department than the right person.
If a customer wants to talk to someone in technical support, the last thing they want is to contact general enquiries and have to be re-routed. Even worse, if they don’t know what department they need to talk to, they could end up being passed around until they find the right place.
Until now, call routing has often been done through IVR software. (Yeah, we cringed too.) IVR is infamous for its problems and how frustrating it can be for customers. Brian Neilson, director at BMI-TechKnowledge, says that IVR "can easily become one of the most annoying and frustrating interfaces known to man." And software like IVR is linear. It requires a clear line to follow, meaning if a customer changes their mind, they have to start all over again. And that can add needless minutes onto an interaction.
This is where AI once again saves the day. Chatbots can be taught to understand intent, rather than simply recognising keywords. This makes it possible for them to identify commands from a normal conversation. So if a customer changes their mind, the AI can pick up on that too and change its implementation accordingly. Much cleaner and much faster.
And customers can simply ask for a department and be routed directly to it. With Gnatta, the AI does exactly that. So if a customer were to ask to speak to ‘returns’, Gnatta would automatically route the interaction to that team where someone is waiting to answer the customer’s question. No more passing a call around departments and no more pointless interactions that skew your score data.
It's pretty clear that AI-assisted agents are becoming more and more complex and more and more intelligent. The newest technology is really pushing boundaries that simply make everything about your customer service that much better. The benefits of customer service automation are always ongoing. That's exactly why we use an AI-assisted agent at FM – we're all about fantastic customer service. If you'd like to learn more, sign up to our newsletter below, or even contact us – our AI will be with you shortly.