Discovering Smart Systems in 2017
How can brands make use of AI and bots this year?
This post is the second in a series of three, looking at the prevailing tech trends that brands should be thinking about as we begin a new year. The first post looked at New Realities. An introductory previous post pondered why these trends seem to be same as last time round.
Along with Virtual Reality, last year saw the re-emergence of another technology that has been touted for decades but this time is serious: Artificial Intelligence. AI, which in my view needs a bit of a rebrand to shake it’s ’90s connotations (hence the trend of ‘Smart Systems’), is now supported by enough computing power and memory to flex its muscles and experiment with a particular strand of AI that is responsible for much of the fresh excitement around it. That strand is Machine Learning.
It was no surprise that Google were at the centre of the most headline-grabbing achievement for AI and Machine Learning last year when it’s subsidiary DeepMind tested it’s AI skills against the world’s top Go players, and won. Elsewhere, self-driving cars — a mixture of AI software, sensors, and of course mechanical parts — received plenty of attention last year, and kicked off this year with more of the same as CES. But what does all of this big thinking meaning for brands?
Well there is certainly a huge, ongoing story around improved efficiency — and the changes this efficiency will lead to — brought on by AI. This was again highlighted by a DeepMind achievement — this time by significantly improving energy efficiency at Google’s huge data centres. The behind-the-scenes benefits of AI are already being used in-particular by many tech companies and will cross-over to businesses of all shape and size. This probably sits in the realm of operations rather than marketing however.
In the marketing world, current conversations around ‘customer-facing’ AI quickly get round to ‘bots’. Bots, or chatbots, are AI-powered conversational software capable of conducting human-like conversations with customers, most likely, but not exclusively, using text. This software has been around for a few years but the big news as of late-2016 is that they are available on, and being supported by, Facebook Messenger, a platform with 32 million monthly users in the UK alone. This fusion of technology and platform presents a fascinating opportunity for brands to have meaningful ongoing dialogue with customers on a platform that many users have migrated to for everyday communication.
To capitalise on this opportunity, I believe brands need to think about something I discussed on a previous post: how to deliver a customer experience that is quicker, better, or different (in a positive way of course) to the communications channels already open for customers to interact with brands. Ultimately, customers have little to no interest in the technicalities of how a brand service — in this case a form of customer service — works, however ‘smart’ it is. Their concern is for the output; the experience they have. The challenge for brand bots then is to provide a useful, even delightful, experience for the customer that is somehow better than (or augments) what is already on offer, whether that’s face-to-face conversations in a store, an enquiry on the phone, or via email or online chat. All of these options are interfaces to have dialogue with real people, who yes, can be useless sometimes, but at other times can be pretty damn fantastic. My best, and most memorable, customer service experience last year was a ten-email long conversation with a customer services assistant at a US-based clothing company who showed astounding understanding, empathy, enthusiasm, and patience, as I faced the common annual challenge of getting my partner’s Christmas present spot-on. A bot would have struggled to help me in the same manner.
But the excitement around bots is very much valid if you think about what they have the potential to do. The combination of smart algorithms, endless data sources (including about the customer), and less-than-a-second response times, suggests the potential for customer service experiences that go far beyond anything users have seen before — not to mention the huge plus point from a business perspective of these efficiently running 24/7 and at any scale required.
Whilst for me the real value in bots will therefore be found in these quicker, better, and always-on (or always-available) experiences, there is value to be had for brands in first dipping their toe in the water with bots that are perhaps at first a little less ambitious. Maybe this is a bot with a very single-minded function, or one that takes existing functionality (perhaps within a brand’s app) into the conversational environment of Messenger. This is what I mean by a bot being ‘different’ — taking something that can be achieved elsewhere, such as making a hotel booking, into the messenger environment — which in turn, should aim to make the process both ‘quicker’ and ‘better’ for the customer. For the brand, this bit-by-bit approach will both get them into the messenger environment, where so many users now are, and will give them a chance to tackle one of the challenges of this opportunity: finding their conversational voice in an environment where users’ wants and needs may be articulated in many different ways.
What should brands do this year?
- Find out how your business is using machine learning to improve operations and understand how this may affect everything across the company (or not).
- Test the Messenger platform with a single-minded bot, but pay careful attention to designing a user experience fit for the medium.
- Start designing new, better customer experiences for your brand that are powered or augmented by conversational bots.