We’re at the cusp of a sharp rise in devices that have no screen but do have conversational voice controls, such as the Amazon Echo. Smart home and Internet-of-things (IoT) objects that respond to users’ voices will improve and become more intuitive with further iterations and wider adoption.
Already they can, for example, dim the lights in a room and play a favorite song. With practice, and, by the virtues of machine learning, these user experiences will become ever more intuitive, capable, and innate.
Beyond the IoT, brands are seeing bots as a new type of media – one that can be harnessed to expand a company’s reach to new customers and networks. As brands use bots more and more to handle customer interactions, make recommendations, and help fulfill requests, those brands and the bots themselves also absorb insights about customer behaviors and needs.
Bots have the important ability to share common context, so they understand when a user is continuing to speak about a topic, or switching to a different one. For example, the user can say, “What will the weather be in Santa Cruz this weekend,” and then “Ok, book a hotel there,” and then “Would be nice if you could order flowers for my wife to have on arrival” — and the bot will access the appropriate services, sparing the user from launching separate apps (and starting over from the beginning each time).
Through the ongoing study of conversations and dialogue patterns, bots will advance to better anticipate customer desires (perhaps even before the customer is aware of them). And as machine learning helps to replace hand-made algorithms with models that are generated statistically, we’ll likely see a reduction in the number of engineers writing code but an increase in those tasked with curating machine learning models and designing even more successful conversational user experiences.
These conversational interfaces won’t just replace much of the functionality of consumer apps; they’ll also impact business applications. One intriguing aspect of the knowledge coming from ongoing bot use is what will happen when conversational artificial intelligence is combined with data analytics. Analysts may be mired in a flood of raw data – a daunting challenge to fully interpret using just traditional tools like spreadsheets and manual analytical queries. However, by bringing the power of conversational UX over to the B2B side as well, analysts can speak or type to an AI interface integrated with their data analytics platform.
Bots are also helpful to analysts because they can offer insights into raw user requests, getting to the heart of what users are actually interested in (as opposed to traditional analytics showing which buttons they click and how much time they spend on a page). This expedites what can be gleaned from users and speeds up improvements not just to customer experiences, but brand-wide.
Imagine an analyst freed from spreadsheets and simply asking aloud, “What was our most popular item purchased by women between ages 18-36 in the last three months? Are there any common trends such as geo-location, or time of purchase, that are notable?” (In fact, Statsbot has created a B2B bot integrating API.AI, Slack, Google Analytics, New Relic, and Mixpanel to serve use cases just like this.) Receiving instant feedback to such queries is as valuable to the efficiency of internal company business as it is to enhancing experiences for customers.
These conversational experiences become all the more compelling when the AI component advances beyond simple listening and task completion and actually serves as a kind of collaborator, intelligently offering services or answers that the consumer could find interesting. As these AIs learn, a future where they act as the faces of brands, as personal assistants, and as valuable co-workers, holds a heck of a lot of promise for all involved.
Ilya Gelfenbeyn is CEO of API.AI, a provider of conversational user interfaces.