Facebook says it will soon be stopping work on ‘M’, its virtual assistant that helped users make dinner reservations or get gift suggestions. What does this mean for the use of chatbots by other brands?
It’s actually a bit more complicated than that – users still launch chatbots for customer service using the base M technology, and its key-word driven suggestion feature will stay in place. But at the same time, M had some issues – it never got out of beta and only ever had a few thousand users. If Zuckerberg cannot succeed with chatbots, does that mean the rest of us should be cautious?
The lesson from the Facebook episode should be that even today’s best algorithms are a long way from being able to really understand all the nuances of natural language. Brands need to use virtual assistants and chatbots for a narrower, simpler, more circumscribed set of tasks so that the benefits of automation can then be reaped.
Facebook made a classic chatbot error by giving M tasks that were far beyond the capabilities of machine learning technology, thus it had to employ an army of real staff to take over when M could no longer help. Too many tasks it managed ended up not being automated, so it ended up being an unsustainable cost centre.
The good news is a large number of organisations have already appreciated the proper role of the technology, and are making them a successful part of their DNA. Organisations are increasingly using chatbots in their business strategy from booking tickets, ordering food and make restaurant reservations, helping to improve customers’ credit scores, acting as a personal shopping assistant, personal PA, teaching/educating – even dispensing medical advice and legal services.
Bots – the new app?
And as more compelling use cases are appearing showcasing the wide range of chatbot services out there, the IT industry is getting excited, too. “Bots are the new apps,” declared Microsoft CEO Satya Nadella in 2016, while for Gartner, by 2021, more than 50% of enterprises will spend more per annum on bots and chatbot creation than traditional mobile app development.
This is also the feedback we received at a recent practitioners’ event we organised in collaboration with the UK Contact Centre Forum. Attendees were drawn from a diverse range of backgrounds, numbering senior UK customer contact practitioners from the legal, retail, health and manufacturing industries, together with technology vendors and consulting firms.
And from the get go, it was clear that Artificial Intelligence (AI) is high on brands’ radars, with the discussion focusing on what it will soon mean for UK brands. The consensus reached: most contact centre managers see chatbots as a useful CX tool – but most are still wary about how this new technology can be seamlessly integrated into the customer experience.
Attendees at the roundtable were also unified in their belief that mobiles are now imperative in winning customer attention, stating that Messenger and other IM platforms are where customers want to talk. After all, customers are no-longer willing to spend ages trawling for results – they want everything in one easy to access place, such as WhatsApp.
That has big implications for chatbot usage. As the M issue shows, how much can they be used effectively – and what should brands do with their in-house human resources as a result?
Some of the practitioners at the roundtable felt AI would be a great opportunity for organisations to differentiate, providing enhanced customer engagement for more complex issues via real human agents. While some attendees were adamant that customers prefer bots for simple tasks, others did not feel this was the case. Some noted that although some customers were tech-savvy and feel very comfortable dealing with bots, others had an aversion to virtual systems, always preferring to interact with a human being. This was especially so with transactions where customers expect human contact, such as when dealing with sensitive and highly personal issues, such as their financial transactions.
So UK organisations are increasingly turning towards conversational chatbots for their messaging apps. The skill sets bots are equipped with should not be over estimated, as in the case of M – but nonetheless, they are developing all the time, and are a definitely a cost-saver when tightly integrated into the business and used for simple automated tasks.
Importantly, attendees at the roundtable concluded that for chatbots to be a success, customers must always be offered a choice. However human-like bots become, if the user prefers to interact in a different way, or over a different technology medium, then this has to be factored into any firm’s outreach strategy, as it’s clearly omnichannel in action.
By taking this route, brands can ensure chatbots really do enhance customer service, save customer time, automate repetitive work – and help your bottom line.