One of the biggest challenges facing contact centres is agent efficiency. Going hand-in-hand with agent performance pressure is the inevitable agent burnout which leads to agent retention headaches for collection agencies. While ever tightening budgets drive contact centres to try do more with fewer recourses, this only exacerbates the problem. In view of this, the high agent turnover is not surprising, but what can you do to remedy this?
The answer is simple – automation.
With automation, the saying “work smarter not harder” applies perfectly. Automating with conversational AI technology does not mean you send all your debt collection agents packing, it means each works to their own strengths in a partnership, and they complement each other. AI automation excels at handling the routine but time-consuming tasks, whereas a live agent is good at tackling the more complex issues.
Do you want to know how to do more with the same agent resources, and how to achieve 300% agent efficiency with 75% customer conversation automation? Then read on.
In the debt collection world, the dialler is the traditional go-to method of customer engagement. At the moment, diallers – agents on a phone - account for over 70% of debt recovery activities, but one look at the stats will show you that this should change. And thankfully, it is changing.
Figures based on Webio clients’ stats
As these numbers show, the dialler is not efficient. Look at the Outbound Effective Rate which stands at only 2.97%. This is a huge amount of activity for a small return.
Inbound calls are more efficient, however, the challenge to have enough agents on hand remains. Diallers still have their place, but to deliver efficiencies you need to bring in digital messaging.
You can do more with digital and here’s why: digital conversations mean automated conversions.
More and more, companies are finding that their engagement levels are going up as the shift to digital in all areas of people’s lives increases. The bottom line is that people are living in a digital world and they like communicating in this realm.
For debt collection, digital text messaging works well for engaging with customers for a number of reasons:
With conversational AI and Natural Language Understanding (NLU), the bot can learn to understand what a person is actually saying and then direct the person down the right route, for example, to an income and Expenditure form, to talk to a live agent, etc.
Here’s how it works.
For the customer, it makes no difference to them is they are communicating with a chatbot or a live agent as long as the job gets done to their satisfaction.
Structured conversations are more straightforward and can be done quickly with the AI chatbots. Examples are:
ID & Verification
Also, entities can be extracted from these answers using Entity Gathering AI and used in ID&V and other activities.
Unstructured conversations are where the Natural Language Understanding kicks in. The answers are longer and more complex, and the AI bot is looking at the intent behind the language. A customer may say things like:
“I need to get more information on this.”
“Where would I be able to get help for this?”
“I think I need your help please?”
Let’s go a little deeper into the tech behind conversational customer engagement.
What does the person want or want to do?
Entities include pieces of data such as amounts, dates, date-of-birth, etc. The system can recognise these and then stores them in a standardised format.
Reading between the lines
The AI system can dig below the surface to find out what the person is really saying, predicts the outcome of a conversation and then directs it down the best route.
In short, working alongside a conversational AI chatbot is a direct win for a live agent.
Triage to the correct area
The chatbot will triage the conversation from the start and send the customer down the best next-step, either to a live agent or on to another automated activity, like a payment.
As an agent types, the Smart Phrases will offer commonly used answers that the agent simply uses without having to type them in.
The propensity can help the agent determine, for example, how vulnerable a person, if their promise to pay is ‘sticky’ or whether they are in a good place.
Use a bot within live conversations
You can switch from live agent conversations into chatbot conversations and back again as needed as the process is fluid and integrated into a single view.
API integration means an agent can have real-time data at hand that is accurate. Integration has other useful functions, such as automatically entering data into the system that a customer has submitted through a form.
Work of the chatbots
Chatbots thrive in tasks such as:
If a live agent has done what they need to do, they can hand a customer back to the AI chatbot to continue with automated activities. You can see this in action in these success stories.
Watch these videos for real use cases that seem too good to be true at first glance but prove 100% true once you dig into the figures.
The first is a car insurance company that automated actions such as quotes. The second is a credit and loans company that saw its automation figures skyrocket: 77.2% automation; 380-420 daily closed conversations by agent up from 50-70; and after-hours conversations taken care of by the conversational AI chatbots.
But don’t take my word for it, watch the video clips from this webinar for yourselves.
Use Case of a Car Insurance Company
Use Case of a Credit & Loans Company
A good customer experience and positive recovery outcomes are the big wins in credit and collections when using conversational AI automation. If you make it easy for your customers, don't be surprised when they engage better with you and respond better. Real client success stories show that these almost unbelievable figures are possible if you embrace conversational customer engagement using multichannel messaging.
Talk to a conversational AI expert who can show you how automating your debt collection activities will push your engagement rate and response rate metrics to new heights.
This blog is based on a webinar presented by Webio.
Watch it here: