Registration is wrong! Please check form fields.
Publication detail:

There is a firm belief shared by most customer experience specialists that if there is one thing that will remain unchanged in the foreseeable future regardless of disruptive IT technologies it is human-to-human dialogue to address and resolve complex issues. Human dialogue lies at the heart of customer relationship building simply because no automated self-services can fully replace interactive communication between humans today. This is perfectly true, but for many companies it comes at a price, as they need to maintain sizeable call centres, invest in supervision and training of call centre agents, etc.

At the same time, we tend to forget that many automated services we are now using almost daily, like withdrawing cash at the ATM, were once regarded as rather risky or even unthinkable. Trusting a machine to hand out cash, keep records, and conduct money transfers and payments would invite one of those rare banker’s smiles in those days when bankers were wearing sleeve covers. We got used to it quickly, however.

Similarly, talking to a machine today instead of your banker, may be associated with a kind of Star Trek-ish experience (it was one of the first movies, where an excellent example of the voice user interface was once presented to the public). The reality of voice user interfaces (VUI) transforming the customer experience and usage of devices is rapidly evolving however, and the effect is already rather striking.

For example, according to Google at least 55% of teenagers in the US prefer using voice commands for internet search at least once a day. In 2016, over 20% of all Google internet searches were performed by voice (at a rate of up to 1,5 bn voice queries processed per day)[1]. Over 65% of smartphone users used voice assistants in 2015. In China, Baidu’s deep learning based system is already surpassing humans in voice recognition delivering surprising level of accuracy on simple phrases. It should be noted that voice queries are really popular in China because most people find it too complicated to input text manually with a phonetic system for transcribing Mandarin in Latin characters. The same would apply to some other Asian languages.

The biggest obstacle to using VUIs for entering data is accuracy. It can be argued that as soon as the level of accuracy reaches the required threshold in a more or less dependable manner on open, large scale vocabularies, input by voice will start replacing keyboards in most environments.

What does this mean for call centres?

According to McKinsey, advances in artificial intelligence, machine learning, and natural user interfaces (e.g. speech recognition) are making it possible to automate many knowledge work processes that have long been regarded as impossible or impractical for machines to perform. Speech recognition and voice user interfaces are prerequisites for an immense leap in the automation of knowledge work.

As a result of the ongoing advances in speech recognition/ASR, voice biometrics, and speech analytics, call centres in the next 5-10 years will completely transform their operations. Interactions with customers will follow the same workflow regardless of which channel is used, and omni-channel interactions will become the new standard.

Any questions will be communicated via any channel in the most convenient way for the customer with voice interaction supported by each channel. Seamless switch of the communication via any channel will be allowed at any stage of the communication process.

In the vast majority of cases the customers will talk with the machine. The following technologies and approaches will be used:

  • Semantic tagging and inference solution predicting the topic of conversation;
  • Speech-to-text/ASR detecting key issues and picking up the details to allow automated form filling;
  • Speaker verification confirming the customer identity without any passwords;
  • Chat-bot finding the correct answer in a knowledge base;
  • Text-to-speech delivering the answer in almost indistinguishably human manner;
  • Only new topics and highly sophisticated questions will be steered to the human experts;
  • Customer satisfaction level will improve as most of the issues will be resolved much quicker;
  • Sentiment analysis will be used for evaluating the customer satisfaction automatically during and after all calls;
  • The results will be used in a continuous machines learning process.

Sounds incredible? Well, nobody knows the future and one can never be sure what direction the market development will take, but we can see it coming very soon based on the observation of the current trends, and it makes business sense to start preparing today.

[1] OMG! Mobile voice survey reveals teens love to talk