With the number of businesses adopting AI growing by 270% in four years most senior professionals, in the credit and collections industry, will be aware of AI and automation.

When looking at what challenges AI and automation are solving, we can look at the calculations used in the decision-making processes, such as when to reach out to a customer, how much to ask for, what channel to communicate over, and what may be improved by specific AI or automation interventions.

AI-driven customer engagement is all about finding out which customers should be contacted, thus driving down outbound contact attempts, and increasing overall contact rate performance.

So where can an organisation more efficiently deploy its resources for customer conversations? There are multiple answers here, but just a few will include:

• Identifying conversational themes earlier in the process – over e-mail, SMS, Webchat, Messenger, webchat, and telephone-based interactions.
    
• More efficient routing of these themed conversations by either automating or being handled by agents.
    
• Giving agents ‘generated content’ to share back to the customer.
    
• Helping agents to navigate the customer’s concerns using prompts.

So, an organisation can now ask in what parts of the customer journey and decision-making process can AI and automation make an impact? Maybe getting customers to respond to messages through better copy, giving customers better decisions through better conversational design, or using AI to identify the likelihood that a customer will complete a form.

Modernising debt collection

AI and automation are now modernising debt collection through AI-driven customer self-service, which enables more customers to self-heal their accounts.

Anything that can drive that number up is a good thing. Also using, AI chatbots to drive customer contact is first finding out which customers should be contacted, thus driving down outbound contact attempts, and increasing overall contact-rate performance.

AI can help customers stay on plan and AI-powered decisioning, supported by workflows, gives end users much more flexibility. Rather than just “conforming to plan,” it is possible to personalise a plan on the fly, again, within rules.

Personalisation

In collections, the one-to-one personalisation of each customer journey is the goal. With AI, the personalisation of customer journeys is more easily developed and delivered. Identifying entities and intents early in conversations helps you understand what a conversation is trying to achieve.

Being able to offer conversational engagement that gives personalised options to customers based on their account status, or customer personae enables you to deliver services that feel custom developed to their needs. And being able to understand how a conversation is progressing and how likely it is to achieve its goal is a valuable AI outcome.

If a person does not fill in their income- and-expenditure form, can we redesign the flow of these questions, can we prepare the customer better, can we follow up in stages to collect all the desired information?

So, using AI and automation to connect a customer and an agent is a perfect example of vulnerability being identified early in a digital conversation and the business proactively creating a different action plan for every circumstance.

Highly personalised customer journeys are based on customer-specific circumstances is possible with AI, and with our Propensity Studio, we are changing the way collections teams interact with their customers. Conversations can be personalised at a point in the conversation based upon what the customer response is. So no more, one fits all approach.

AI in customer engagement is all about your choice of words

AI-driven customer engagement is all about finding out which customers should be contacted at what time and in what manner both from a channel perspective and the actual words used to increase overall contact and engagement performance.

The actual words you use is an area that isn’t given enough attention in my view. From experience, getting this right will mean your message will resonate with your customers in their current circumstances fostering that willingness and desire to engage with you.

How to deploy AI in customer engagement

I wouldn’t recommend you jump in at the deep end and rush into deploying AI into your collections operations. But as with any aspect of new technology deployed in any part of the business, including collections, best-practice concepts have been developed, and when followed ensure a higher level of success is achieved. 

Here are some that have worked for clients that we have engaged with:

• Start simple and slowly develop towards complexity.
   
• Work with a technology partner you trust and who understands your business.
   
• Build/Support an internal champion and their capability to make decisions in this area.
   
• Work on specific outcomes that are measurable and understand the before and after state.
   
• Think about AI in tandem with your data strategy, what data do you already have that would be useful.
   
• Understand the appropriate risk and de-risk your projects by slowly building up the size of customer groups you're engaging with over a period of time with regular reviews.

Persuade me that AI works

AI isn’t different to any other technology, there will be a certain amount of scepticism and caution about using it and what benefits it will deliver. The classic rules do apply here about new technologies invariably over-promising at their launch with results and adoption being underwhelming in the initial periods.

It’s fair to say we passed this initial period and the results that are being delivered to all businesses, including ones in the collections industry, are real, tangible and significant, which you can’t afford to ignore any longer.

AI isn’t magic, it’s just another technology that you’re using a lot today already. So why not bring it into your business?

• If you use Google search you are using AI.
   
• When you show your smartphone and Apple Pay “pings with paid” you have used multiple forms of ML/AI including facial recognition in a payments process.
  
  
• When you go to a website and use webchat and ask it “where is my order” and receive details in the back and forth, that is conversational AI.

So, should you let AI manage your debt-collection process?

In truth, no. Rather, you should let it augment your collections process. You should apply the AI and automation that has a “high probability of being successful” before trying other forms of AI (early wins will build confidence) but be sure that AI can definitely help you with processes.

AI and automation are tools that if put to their best use can really have significant impacts on not only the operational side of the debt collection process but also the customer relationship.

In this environment of rising consumer debt levels and people exhibiting vulnerability, those businesses can shift the focus from being processes-driven to a customer-first approach using technology to predict conversation outcomes and guide customers on highly personalised journeys that are specific to their circumstances. 

AI and automation have permeated the collections space, they are here, they are working. So starting to think about how your business can use them and ensure that you do not get left behind.

One important metric I will leave with you is that our clients typically double the number of processes they apply our technology to within the first year of operations. Because it works.