Running a contact centre is expensive. In 2021, 21% of industry professionals perceived their contact centre as a ‘cost centre’, and 11% even went so far as to describe it as a ‘necessary evil’. In recent years, companies in all sectors have been trying to reduce the number of customer service calls made by customers. Knowledge bases, livechat, and FAQs have now become standard across the customer service industry. And some businesses have taken the extreme measure of burying their customer service phone number deep within their website to deter all but the most urgent of calls. In some cases, businesses are using AI automation to deflect customers away from phone calls and towards text-based chatbots. Such bots are usually offered over social media or instant messaging channels.
However, these initiatives have all had limited effectiveness. Only 6.9% of contact centre interactions take place over live chat, while phone calls (both inbound and outbound) still account for 68.5% of all interactions. To put it simply, contact centre staff still spend the majority of their time answering phone calls, and this can have significant costs.
Recent research by Gartner indicates that contact centres are still spending heavily on AI tech. Worldwide spending on conversational AI solutions for contact centres is predicted to reach $1.99 billion this year. Today just 1.6% of interactions are automated using AI, but Gartner estimates that 1 in 10 agent interactions will be automated by 2026. These investments could save contact centres a substantial $80 billion, largely in reduced labour costs. Text-based solutions are only capable of deflecting a minority of user requests, firstly because they tend to be simple solutions that constrain the user, and secondly because people naturally reach for the phone.
So what accounts for the enormous cost-saving potential? The answer is the forthcoming rapid growth of conversational AI-powered voice assistants. In contrast to chatbots, these assistants have greatly improved support for voice interactions and underlying semantic understanding. They take advantage of the latest advances in natural language processing, automatic speech recognition, and machine learning to hold extended, engaging conversations with customers.
Why text-based solutions are not enough
It is easy to understand the appeal of no and low-code platforms which promise a DIY solution that is easy to build and fast to deploy. However, banking clients will quickly discover that these services have considerable limitations that make them unsuitable when scaled.
Firstly, these services often lack sophistication and are only capable of answering extremely basic customer questions. Often the logic and NLP recognition that underpins these systems is extremely rigid, and they are heavily dependent on a series of rules or keywords. Even minor deviations in how you phrase your questions can cause the bot to misunderstand or return an error. As a result, you will often need to repeat yourself, rephrasing the same question again and again until you find the magic keyword that kicks the chatbot into gear. Additionally, if you speak outside of the bot’s remit, it can spiral in unexpected and confusing directions.
Customer service requests in banking are often complex and can be laden with financial terminology. Simply put, the chatbots that no-code platforms power are not sophisticated enough to meet the needs of banking customers and fall far short of their expectations.
Secondly, these no-code bots are usually built primarily around text-based instant messaging channels and treat voice support as secondary. This means that the speech recognition and natural language processing capabilities of these services are very limited when these two layers should be working strongly in unison. Voice assistants are also often unable to accommodate interruptions or background noise. They might work well in laboratory-like conditions, but they quickly begin to struggle when faced with everyday customer service realities of fragmented speech, poor call quality, and ambiguous requests.
Ultimately, the problem with weak chatbots and voice assistants is that you need to make your customers change their behaviour in order to use them. As mentioned above, the majority of customers are still calling contact centres and not making use of livechat and chatbots. Customers need to be met where they naturally reach out and banks need to invest in voice automation as well as texting.
The benefits of supporting voice
Thanks to advances in machine learning, natural language processing, automatic speech recognitio, and speech synthesis, virtual assistants are now capable of holding intuitive and engaging conversations with customers over the telephone. With these next-generation assistants, call centers of the past are far more efficient than they used to be with IVR and DMTF systems.
Conversational AI assistants can retain context across conversational turns, allowing customers to offer information in the order that feels most natural to them. This makes the bot’s conversations with humans feel far more natural and fluid as the customer is gently guided towards the accomplishment of their goals. With superior recognition of meaning, conversational AI services understand customers even when they hesitate or interrupt, and background noise is accounted for.
For financial organisations of all sizes, there are benefits to providing conversational AI automation over the phone:
1. Always available
Day or night, conversational AI assistants can be ready to assist your customers, even when your contact centre is closed. No virtual assistant ever has to take a day off, meaning you can finally deliver 24/7 customer service automation at scale.
2. Immediate answers
Forget waiting on hold. You simply need to dial in and you can access instant customer support. 23% of customers abandon a call after waiting for less than 29 seconds. Conversational AI helps your bank retain these customers and improve their loyalty by attending to their needs faster.
3. Intelligent conversations
With the ability to retain context and decipher customer meaning even through background noise, conversational agents can hold human-like conversations. The result is that interactions feel intuitive and engaging. When talking with a virtual assistant, you’ll never need to repeat yourself, and you’ll always be understood however you choose to express yourself.
4. Improve accessibility
Typically, we can speak faster than we can type. Conversational AI improves the speed and accessibility of online banking services by allowing customers to phrase their requests in natural language.
5. Reduce costs
As outlined earlier, conversational AI technology has an extraordinary cost-saving potential. With the ability to automate a large proportion of common customer service calls, conversational AI speeds up call resolution time and should significantly reduce labour costs.
The voice assistants that action.ai develops for banks are domain specific. This means that they can more accurately answer a wider variety of everyday banking tasks and requests. Our goal is to deliver the next generation of conversational AI technology and to provide exceptional customer experience automation.
