subscribe: Daily Newsletter

 

How chatbots will evolve into practical tools

0 comments

With research showing that the number of new app downloads is slowing down, it’s clear that customers are becoming pickier about the apps on their phones and tablets. The slowdown also points to a phenomenon of ‘app fatigue’ – we’re getting tired of continually updating all our apps, of learning different UX and navigation principles, or remembering passwords.
There’s a theory – popularised by a book, a talk at the South by SouthWest conference, and the hashtag #NoUI – which says the best user interface is, simply, no user interface. #NoUI tries to address the issue of app fatigue.
Craig Heckrath, head of products at Intervate, explains: “In the spirit of #NoUI, many of us are enchanted by the idea of a more natural, more human interface that could replace the small rectangle of smartphone glass which seems to continually demand our attention. And it’s for this reason that the idea of chatbots holds such universal appeal: verbal language is perhaps the most natural, most human way of communicating.”
Just think of the apps that are most often used. For most people, chat messaging apps like Whatsapp, Facebook Messenger and WeChat reign supreme. Many analysts predict that these platforms will develop into whole new ecosystems of services – the likes of ecommerce, payments, personal concierge, gaming, and more.

The promise
Intervate head Peter Reid notes that chatbots in the workplace perfectly combine the popularity of messaging as an interface, with rapid advancements in the field of artificial intelligence. They promise to help customers, staff members, partners or other parties engage more fluidly with an organisation.
“Chatbots are often hailed as the end of call centre frustrations – no more waiting on line, being transferred to other agents, or getting cut-off. Instead, customers can chat at any time of the day or night, with intelligent bots that can interpret various styles of language (in fact, various languages), and provide useful responses,” he explains.
For the organisation, chatbots can be used to automate certain repetitive tasks and common queries – such as IT helpdesks resetting users’ passwords, for instance. It’s a way to automate low-value activities, and redeploy the human staff to focus on more strategic work.
But these are the grand promises of chatbots. So far, the early examples of bots are fairly limited in their scope, and can become quite frustrating if users try to perform more complex tasks – above and beyond what the chatbot has been specifically ‘trained’ to do.

Cutting through the hype, getting down to business
The coming months and years will see businesses becoming more realistic with the way they implement chatbots into the organisation, notes Heckrath. “The focus will be on integrating the chatbot channel into the core systems and data repositories of the organisation. Those that try to become too ambitious too early will probably struggle to get their chatbot programmes off the ground.
To increase the chance of success, Intervate advocates a five-step process to getting started on the chatbot journey”:
* Codify those common queries, and commonly-used information sources that could be used by a chatbot, to offer immediate value to users. Look at those repetitive tasks that are easy to automate and to surface in the chat format.
* Begin with a few limited channels – for some organisations, this may be Skype for Business, or a web chat window. You don’t have to give your bot a presence of every possible channel immediately.
* Monitor the early conversations at a very granular level of detail, use those learnings to understand user sentiment and frustrations, and adapt the bot as quickly as possible.
* Allow your chatbot to “fail gracefully” – if it doesn’t know the answer to a question, allow it to become a simple digital assistant, which helps to patch the user through to a call centre, or direct them to a support email address, or escalate to a particular individual.
* Don’t oversell the technology, to your stakeholders or your users. Don’t try to pretend that your bot is a warm-bodied human being. It’s important to set the users’ expectations up-front. Generally, people are quite forgiving if your bot has an open and honest personality.

Glimpse into the future
By getting started early on a chatbot journey, companies will be perfectly poised to benefit from the huge strides that developers are making in refining natural language processing, enhancing the sophistication of artificial intelligence, and creating ecosystems where specialist bots can communicate and hand-over to other bots of a different specialisation.
Reid adds: “As chatbot technology continues to mature, we’ll start moving into the realm of predictive algorithms – where the chatbot can start initiating the conversation with the user, based on predictions about what the users will need. This has the potential to dramatically enhance customer service and satisfaction levels. In time, your chatbot will be able to synthesise and crowdsource information from a vast number of information services, to provide the best-possible engagement with users.
“From there, the future is unknown. In much the same way that we could not have predicted the rise of apps like Uber and AirBNB a decade ago, it’s virtually impossible to imagine the new opportunities that chatbots will create in the future. Perhaps we’ll see entirely new businesses growing out of the chatbot revolution?”
Heckrath concludes: “Whatever the future holds, chatbots will play a role. There’s never been a better time to explore how your chatbot could help spur your company’s digital evolution.”