Bots: artificial intelligence in dialog with customers
Artificial intelligence and machine learning – and the impact of these technologies on corporate success – were the subject of a study we recently published in partnership with Crisp Research and Hewlett Packard Enterprise (HPE). In this study, we explain the “traditional” applications for these technologies in a commercial context, and take a look ahead into the future – more specifically, at how companies interact with their customers automatically and digitally via speech control and digital agents.
Conversational interfaces – bots and intelligent assistants
Efforts to optimize the interface between man and machine are nothing new; work has been ongoing in this field for many years. In our increasingly mobile world, the act of entering data using a keyboard and mouse has become obsolete. Touch-screen displays serve only as proof of this shift. A new era has begun, and conversation-based interfaces are on the rise.
Bots, or intelligent assistants, have a key role to play in this process. These systems and products rely on machine learning technologies – particularly speech recognition, natural language processing and tools capable of analyzing gestures, expressions and faces.
So, in our evolving online world, digital products and services will no longer be able to survive without these new interfaces to interact with people. This is true across the entire spectrum of digital communication – whether a customer is interacting with his online banking system, controlling his smart home devices, or communicating with the customer service department of a public transport company. Products like the Amazon Echo, Google Home and Siri are clear indicators of the trend towards the further commercialization and “productization” of machine learning technologies. Many developers and companies are also working on their own systems to equip their products with digital assistant functionality.
What role do digital/intelligent assistants play in the machine learning strategy of the companies surveyed? 46 percent are evaluating the use of these technologies within their own companies. 24 percent are already at a more advanced stage, and are rolling out projects and conducting tests. Just five percent of companies are actually using these technologies in their day-to-day operations – an innovative approach at this early stage of system development, at a time when some platforms (such as Siri by Apple, or Chatfuel by Facebook) are still being rolled out for open use at a painstakingly slow rate. Many companies are, of course, developing their own systems and digital assistants, although this strategy requires a significantly greater investment. This investment may well be one of the reasons why 26 percent of companies have not yet made any move towards implementing assistants. Only when bot platforms from major Internet and cloud providers become fully open, and when other bot technologies are refined, will companies be able to adapt the bots to suit their own needs.
Even today, at this early stage, it is clear that the use of intelligent assistant systems not only holds great promise in terms of innovation and automation potential, but also comes coupled with significant risk. If the systems are not sufficiently sophisticated, or if the models are poorly designed and trained, companies run the risk of a long-term loss of customer confidence and damage to their reputation. After all, these systems interact directly with customers – and customers rarely tolerate mistakes.