The panel unpacked topical issues such as “What are the best strategies?” and “What is the best mix between people and technology?”
Nick Cherry from Phillips & Cohen and Frank Sherlock from CallMiner joined Mark on the panel, with Chris Warburton, Director at ROStrategy, hosting the discussion.
♦♦ Here are some highlights from the lively discussion. ♦♦
As we all know, Covid caused a stir in many aspects of our lives. One of the areas that got hit hard is living costs, which triggered the cost-of-living crisis. This resulted in an increase in the number of customers struggling with debt. This, in turn, has put pressure on the collections industry to keep up with the demand and stay ahead.
Covid also forced the credit and collections industry to rethink its approach to customer engagement. Instead of in-person conversations, automated debt collection has stepped into the spotlight. It's like Covid hit the accelerator pedal for the industry to turn automation from a mere 'to do' list item into a reality.
Throughout this intense period, people have found it harder to have conversations around debt, which has made it challenging for companies to reach out to their customers in difficult situations.
To handle the increase in customers who need to engage about their debt, especially if they are in a vulnerable situation, we need to take a more nuanced approach, for example, is a digital or human connection the best option?
So, when it comes to technology, it's important to make sure that all the different channels work together seamlessly. That's where digital technology comes in to make things easier for everyone involved - the agents and the customers.
Thanks to the advances in API capabilities, agents now have access to real-time information that makes their job easier, which in turn makes the customer's experience much better. Plus, the AI chatbots get accurate information about the customer, so they can provide superior service.
What’s more, when it comes to conversational design and using different channels, flexibility is key. You don't want to get stuck in a rigid approach where you only do things one way. Each customer is unique and has different needs, so it's important to be adaptable and not pigeonhole customers. Find what works best for them in any given situation.
One of the challenges is learning to engage empathetically with people who are really struggling financially and find themselves in debt for the first time in their lives. Customers in mounting debt are in a fearful place, and this makes them put their heads in the sand and not want to connect with their creditors.
In these cases, communicating via digital channels helps them as there isn’t the ‘embarrassment factor’ when talking with chatbots. Digital is also good at handling transactional tasks and helping people manage their debt.
The concern is, though, can we still be empathetic towards vulnerable customers if we go digital? And when do you ’flip the switch’ from live agents to digital? The consensus on the panel was that you will always need to have human agents available to talk to people who need a more nuanced and complex issue solved.
Those that can’t, won’t or just don’t like to engage digitally need to have the option of speaking with a live agent.
It is important to get the balance right between human agents and chatbots. There is a strong cohort of people (30%-40%) who will always prefer not to talk to a live agent about their debt. Webio’s 2022 consumer report found that although the 55+ year-olds favour emails and phone calls, there was only a 7% gap from the 25-45 year-olds who prefer digital channels.
In other words, you cannot stereotype how each age group would like to communicate with businesses, and the older generation is getting more tech savvy all the time.
Contact centres have found that after adopting conversational AI automation to handle the more straightforward queries, the live agent conversations tended to be longer. One customer example showed there was a drop of 30% in customer calls, but at the same time, an increase in call length.
This is good as it shows the premise is working – let the chatbots handle most of the activity to allow agents the time to deal properly with the more complex tasks. Also, the tech helps the agents to do a better job when on a call (or text conversation) and this helps them feel they are adding value.
A big win is that agents who feel valued will more likely stay and this helps with agent retention.
The panel agreed that you’ll always need humans as part of the process, and we have yet to see a chatbot that can truly show empathy – they can only mimic it.
Today, we have whole generations that grew up in a digital world with smart phones never far from reach. Even the older generations are becoming more and more digital-first and expect to be able to contact companies via mobile apps and omni-channel communication avenues.
This multichannel approach requires good orchestration software and analytics to understand how customers engage across their journey. Including clear signposting and offramps into your conversational design and self-service portals is important - customers need to know they can call on an agent whenever they need to and not feel frustrated and stuck with no answers to their queries.
There is a large volume of data and user analytics to understand where the technology is working or failing. You can look into the data and then go back and correct what’s not working, make the modifications and feed it back into the system. Analytics helps you to refine the customer journey for the best outcome.
A good approach is to design the automated conversations with what you know, and then use the analytics data to make it better. This iteration is a continual process that helps get the job done better each time.
The key part of the conversational analytics is to link the data together. For example, you can track how someone starts in webchat and then breaks out into a conversation with a live agent. This shows you how far a chatbot can go before a human must step in.
See: What is ChatGPT all About?
ChatGPT is compelling and intriguing technology that brings many questions to the fore. How will it change the customer engagement dynamics? Where is this tech taking us? Can it ever replace humans for in-depth conversations?
The panel's view on ChatGPT was that it is powerful, but not yet ready to tackle the really complex interactions with customers. It is an excellent tool that has more functionalities beyond ‘chat’, e.g. it can be used in conversational design, improve speech analysis, summaries, transcriptions, etc. but it’s not at ‘prime time’ stage - yet. It can be a great employee co-pilot to take the pressure off employees and do things quicker, however, it’s not accurate enough yet to be fully trusted.
The panel predicted that in 4 to 5 years' time, we’ll have a personal digital assistant on our phones that will take ChatGPT a step further.
What stands out about ChatGPT is the brilliant marketing that launched it to the world. It woke a lot of people up to the potential of AI and Machine Learning and how they can be used.
A pressing problem with AI is how to guard against bias? The first step is to be aware of it and to keep training, keep iterating to get it right. Regulations also act as guard rails to keep things in check. (See: Debt Collection Regulations & Compliance: What you Need to Know)
When it comes to creditors and their responsibilities towards their customers, it's important that they provide a clear, easy and safe environment for positive engagement. After all, nobody wants to deal with a complicated and confusing process. And, this shouldn't come at an additional cost or time for either the customer or the company. The idea is to make the customer's experience as smooth as possible.
One measurement to look at is customer effort and you can identify areas for improvement using analytics tools. With technologies like propensity guidance, intent recognition and sentiment analysis, customer interactions can be more personalised which results in a better outcome for the customer.
Propensity engines can predict the outcome of a conversation in real-time, which allows companies to route the conversation down the best path for the customer. This helps to ensure that the customer's needs are met as efficiently as possible.
But, it's not just about measuring metrics like cure rate and cash collections. We also need to identify where customers may be struggling along their journey and find ways to resolve these issues. By understanding the pain points and areas of friction in the digital journey, we can work to remedy the problem for the customer and help them as quickly as possible.
At the end of the day, it’s simple: treat people with respect. We need to get customers into a conversation and how we do that has to be by respecting people and not ordering and intimidating them. This approach will ensure a win-win for both the business and the customer.
Furthermore, customers want to be in control of their own finances and solve their own problems, so digital debt collection is usually well received.
As the wave of change sweeps over us, we might want to catch our breathe, but it seems that innovation is not slowing down. AI, ChatGPT, Bard and the like will affect huge change as they get fine-tuned over the next few years. This will make the digital landscape smoother and more effective. AI is here to stay.
As collections companies continue to embrace technology to improve customer engagement, they need to make sure to keep within the fences as to what is permitted and not have to catch a reality check by taking a step too far.
Also, we will see a continuation of the good progress that is already in full swing, such as the bringing together of different channels into one place and AI chatbots learning at speed, API allowing for more integration of your full tech stack into conversations, and an optimisation of digital alongside humans. The future is bright.
And lastly, thank you to Credit Connect for putting on a great event.