First Contact Resolution - is it the right metric?
2 Aug 2024
6 min read
Is First Contact Resolution the best way to judge AI in customer service?
When ranking the metrics of success in customer service, first contact resolution (FCR) rate comes near the top of most lists. Getting a high FCR score means that you are preventing one of the most frustrating events for both customer and agent: repeat tickets.
Repeat tickets can effectively kill good customer service. Having customers come back time and again means that the number of tickets goes up, meaning more work for your customer service team. Customers who contact you again are rarely more happy than the last time they spoke to you, so you can start to erode a good reputation.
So when it comes to using AI and automation to enhance your customer service, a natural expectation is that it will improve your First Contact Resolution rate. This is not always the case though, but even if it doesn’t, it doesn’t mean that AI is bad. Let us explain.
What is First Contact Resolution?
Quite simply, first contact resolution rate is the proportion of your tickets that are solved the first time a customer contacts you. It’s also known as first call resolution, or first touch resolution. This means that the customer has not had to contact you again, nor have your agents had to investigate and follow up.
How can you improve First Contact Resolution generally?
Leaving AI to one side for a moment, the general way that you can improve first contact resolution is by ensuring that your agents have all the tools at their disposal to be able to solve the majority of cases in one go.
For instance, if a customer wants to change an order that has not been shipped, then your agents need to be able to edit the order while on the phone with the customer. Or, if a customer wishes to return an item, the agent should be able to access a returns portal or platform and issue the customer with a new return label and clear instructions on next steps.
In other cases, it could mean ensuring that the agent has clarity on what the processes and policies are for most common circumstances so that they don’t have to check with a supervisor, or go away and investigate.
Another key factor is that any action the agent takes actually happens. So if an order is amended by the agent, it needs to be picked up by the warehouse team and the change effected. Otherwise the wrong order gets shipped and the customer has to contact them again.
So, ensuring access to systems, clear policies, and making sure that your tech is integrated effectively are all ways you can start to improve FCR.
Is a high FCR ever a bad thing?
Yes and no. Let us explain.
Imagine that your team gets 100 tickets a day. Some of those tickets are customers asking very simple questions that your team can answer without thinking or checking. The sort of questions you get asked every day, like “What waist size are the medium trousers?” or “Do you do Next Day shipping?” and so on.
Most of those questions can be answered first time, thus you get a higher FCR score.
But, a lot of those questions could be answered by having that information clearly displayed on a PDP (Product Detail Page) or a FAQ or shipping page.
If you implement changes that allow customers to find the answers themselves then you might cut the number of tickets you get to 80 a day, but they will be more complex. That means your FCR goes down, but your team has less work, so on balance things are in a better place.
So why might FCR not be a good measure of AI’s effectiveness?
It depends if you measure FCR when a customer reaches a customer service agent, or when a customer chooses to interact with any self-service or AI solution.
If you choose to measure it only when a customer service agent gets involved, then the AI will reduce your FCR by solving the simpler, slam dunk type questions.
If you choose to include any occasion when a customer asks a question, then you may not move the dial at all.
That’s because when deploying a sophisticated AI agent, such as DigitalGenius, you’re looking to replicate what your agents are doing. If you do this successfully, then the AI is just doing what the agent was doing anyway.
The only way you increase FCR is if the AI makes fewer mistakes, or if it can do things that the agent is unable to do.
We believe that an AI should make fewer mistakes than a human. And our solution can do things which agents are unable to such as provide up to date estimates on delivery timelines. However, these may not move the needle that much when it comes to FCR.
Measuring an AI’s FCR rate
Where FCR is helpful in considering AI is when you assess the ability of the AI to resolve issues on the first contact.
To do that, you’ll need to be sure that you know when a customer interacts with your AI and then goes on to speak to an agent. If you are unable to track the whole journey, then you can ascertain how effective the AI is by asking customers if they were able to solve their issue using your AI customer service solution.
On the other hand, if you use an AI agent as a first port of call, but allow customers to move seamlessly to a customer service agent, then you can measure how often the agent has to step in.
So is FCR a good metric when thinking about AI?
The bottom line is that adding an AI is almost certainly not going to improve your First Contact Resolution rate. If you measure FCR only from when the customer speaks to a human agent, then adding AI agents will mean that many of the simpler queries are taken away from agents, meaning they have a higher complexity and therefore a higher likelihood of needing to be solved in two or more contacts.
You could consider it a single contact, if and when a customer is handed over seamlessly from an AI agent to a human agent. In this case, your FCR is likely to remain flat.
There are some ways AI can improve things, by making fewer mistakes, by being able to connect the dots between what customers are asking and what data you have, or by doing things agents cannot do.
One thing is certain though. Adding a sophisticated AI Concierge like DigitalGenius will have many other benefits:
Reduced First Response Time
Reduced Average Handle Time
Fewer Tickets Handled by Agents
Round the Clock Ticket Resolutions
Smaller Backlogs, and Reduced Repeat Contacts
And of course you can use FCR as a measure of how successful the AI agent is. Two different AI solutions could have a very different FCR depending on the approach, so it is still a decent metric to compare different solutions.
If you want to improve your customer service using AI, speak to our team today.
