Provoke series by Mando Agency. Jonathan Seal explores why 2017 is the year of the chatbot and some of the ways in which brands are utilising bots to increase customer satisfaction, improve service and retain talent.
Good evening, and welcome to our Provoke session tonight on The Year of the Bot.
You may be wondering why there’s a picture of a chicken on screen. Well, this is a rooster, and as we’re had Chinese New Year we’re now entering the year of the Rooster. And in Chinese culture the rooster represents two primary characteristics… punctuality and honesty. I think these two factors (being there at the right time, and being authentic) are incredibly relevant for the challenges businesses face today as consumers and employees demand always on response, and the desire to have authentic interactions with the brand. Punctuality. Honesty. And it’s in this context that I think ‘bots make a fascinating and compelling case to be heard in 2017.
But to start let’s go back about 25 years to a quote from Mark Weiser, father to the concept of ubiquitous computing (the idea that computers would just permeate everything that we do)… Think about how smartphones have become utterly embedded into the way we live our lives. We’re at the start that same process for conversational interfaces to digital systems. They’re just going to become normal - boring even. But right now businesses that choose to adopt them can be the leaders and gain real competitive and market advantage. What do I mean?
We know that basic chatbot are becoming an increasingly popular mechanism for customers to interact with brands. Dominos, but also Burger King / 1-800 Flowers / Pizza Hut / Starbucks… : Chat-enabled systems for selecting and customising an order, and track delivery through simple automated messages and responses Uber: This is an interesting one, because you can have a conversation with a friend and then invite Uber to get you a ride within the conversation just by clicking on an address.. RBS and American Express: both have chat bot systems where you can ask for purchase alerts and get supported through service issues like forgotten PIN numbers and lost credit cards. CNN & Guardian: Sends you personalised headlines as messages and you can then ask for more detail on the news that matters to you. So the interesting thing here is that they are being used for sales enablement, core service provision, customer support and a whole host of other things that we’re just beginning to wake up to.
Two nice examples to look at in more detail. KLM have been running a bot for a while now, with three key features. 1. You can get real time flight and itinerary information 2. It actually sends your boarding pass via the messaging interface, and 3. It pushes gate reminders or alerts you if something changes.
And from the sublime to the ridiculous… there’s even a chatbot that toothpaste manufacturer Signal have created that tells an interactive story around brushing teeth so that families get better at it through gamification. So what is it that makes these brands bother to invest in something like this? I mean of course there’s always some advantage in the publicity around novelty, but this is more than that. What is it about ‘bots that makes brands serious about investing in them and adopting them? Well to me there are 4 key benefits.
1. there’s no app to download. The users already have it because app is whatever channel your customers already use. And we know it’s a real barrier to get users to download and keep your app on their devices, especially if they don’t use it every day. All that goes away with conversational interfaces. 2. Secondly bots reduce the demand on users because they don’t need to learn how our app or site works. They just communicate in the way they want to using the channel they are already used to. I don’t need to create instructional videos explaining how to it because it is just SMS or Facebook or Slack. 3. they don’t have to switch context to use your services. That makes it easier for them to engage your services in a seamless way. (e.g. Skyscanner / Uber) 4. Finally the whole experience feels much more like an ongoing interaction, rather than a one-off transaction. The ability to maintain your conversation history and continue from that point seems much more ‘human’ than the normal web experience of “fire and forget”.
Because of this, some people are making pretty bold predictions for how much conversational experiences driven by ‘bots will shake things up… [quote] Now Ted has a vested interest in making this prediction – his company Kik is a messaging platform that includes a bot store. However we do know that companies of all types are bought into the model of investing heavily in creating not only the bots, but the bot frameworks, the tooling, the supporting services and so on to create this ecosystem.
So if we’re looking at the rise of automated bots and a shift towards more conversational experiences, why is this happening now? Well I believe there are 2 key drivers for it…
Even if you ignore bots then messaging apps themselves are incredibly popular. We all know and use a variety of messaging apps like these throughout the day and there are hundreds more. The interesting part of this trend is that they are increasing not only with Gen Z, Millennials and Gen X but also the baby boomer generation who find conversational interfaces easier to process than more graphical and icon driven experiences.
This shift is apparent as we watch interactions of messaging platforms overtake those of social networks. In 2016 the two most popular app downloads were snapchat and messenger, both platforms primarily designed for person to person communication rather than broader social networking.
So one driver is just that we’re already using messaging more and more as a means of interacting with companies and getting things done, just not in an automated way. The second reason is the improvement in AI capability. We’re getting much more comfortable with AI-driven conversation experiences in our homes and everyday experiences. You just have to look at services like Siri and Cortana, and products like Amazon Echo and Google Home to see how pervasive they now are. The fact that the global players (Microsoft, Facebook, Google, Apple, Amazon) are all heavily represented here shows the strategic importance they place in AI and AI-fronted interfaces. And the improvements these systems drive are making it easier to apply intelligence and responsiveness into bots regardless of how we communicate with them. [latency, accuracy of NLP and computer vision]
As a quick aside, this shift has really important consequences for people working in agencies. When we think about designing digital experiences we’re normally thinking of an interaction with some form of screen, where layout, hierarchy of information, design aesthetics, animation and visual feedback are all key elements. However in this new paradigm all those things are fixed as part of the channel you’re using (Facebook Messenger, Slack, Skype and so on). We can’t control those! That means that experience design becomes far more about how you structure and guide people through a conversational flow in order to communicate the brand and tone of voice throughout an interaction. The design process is back to the basics of storytelling and engaging in a dialogue with our customers.
Now I couldn’t talk about chatbots without mentioning the Mando fridge bot. We’ve worked hard as part of our office move to create a more compelling and modern digital workplace. We’ve got more modern tools, we have a robot that welcomes you in our reception (he’s a little sad here because he’d just bumped his head). We have (whisper) Alexa in the kitchen so you can ask it to play music or answer questions…
But a large part of our office looks a little like this… lots of people busy getting on – working on client projects. Now as an agency, we need to keep track of studio time spent on projects for billing, and it’s always been difficult to get accurate and timely data when it’s reliant on people filling in timesheets (which are dull and administrative but important). So we decided about 5 years ago that if everyone completed their weekly timesheets…
..we would all be able to have a nice cold beer on a Friday afternoon! OK, so what has beer got to do with chat bots and artificial intelligence?
We’ll, the beer sits inside a fridge that you can see here with an internet of things enabled magnetic lock (forgive us, we’ve just moved and we haven’t set the lock up properly so if you go up to this fridge it will just open!)
And above the fridge is a screen that displays in real time how much time has been put into the timesheet system.
It also shows names and shames people who are currently blocking the fridge from opening, so that we’re all encouraged to make sure we do get it done on time. But crucially, about 7 months ago we created Mando Fridge bot so you can interact and find out what’s going on.
Here’s a conversation I had with our fridge. I can ask how much of a blocker to the beer fridge unlocking I am, or anyone else is. It also works on mobile devices. In this version I’ve asked it what’s in the fridge, which is stored in a simple database that gets updated each week, meaning I can also ask for my favourite drink to be there and ready for when my timesheets are all in. You can even use the standard voice recognition (also AI) that comes with Siri or OK Google to mean you don’t need to type requests. One thing we did find however was that we had a lot of instances where the bot was reporting to us that it struggled to understand what some people were requesting (see Malibu comment). When we looked into it we found that it was mostly that people were attempting small talk either before or after requesting timesheet data, and the fridge didn’t know how to respond.
Now we’re not trying to get something hat can pass the Turing test, but we spent a little time adding in some basic manners, casual comments and some personality that would go a long way to help make the experience a little less jarring. It’s still along way from perfect but the bot has been in use now for about 7 months and we’re learning new things all the time that help us.
The value we’ve got from this means that we’re committed to extending this into new Mando bots to cover other scenarios, like requesting annual leave, or letting people know you’re sick, or booking meetings within a conversation by inviting the bot to help out…
Here’s an example around booking holidays. [go through demo] Conversational bots can be valuable in lots of ways and we’re using them more and more as part of our core systems , like for informing us of deployments for software builds, or that someone has added items into a Trello board, or that we’re down to work on a task in Jira. For Mando, we see ourselves building up an ecosystem of interconnected and intelligent bots that support our team. Augmenting our people to make them superhuman. Because if you think about it, we’re already all cyborgs. We outsource activities to machines every time we use Google maps or Evernote. This just makes the experience of doing this more seamless and intuitive.
So where do we go from here? Well, if we look at the customer opportunity business case, it’s around [list]
And for employees, it’s around [list]
Some simple examples to explore could be: as an intelligent layer on top of FAQs To enable self-service of low complexity high volume transactions To filter and gain context before customer live chat or calls Improve employee engagement – operating at top of HoN
I hope that’s been useful. Thanks!
Free Speech - 2017 Year of the chatbot
Why should chatbots
matter to my business?
Sorry, the office is now closed.
“The most profound technologies are those
that disappear. They weave themselves into
the fabric of everyday life until they are
indistinguishable from it.”
Mark Weiser, 1991 “The Computer for the 21st
“Chat is going to be the next great operating
system. Apps will come to be thought of as the
new browsers; bots will be the new websites.
This is the beginning of a new internet.”
Ted Livingston – CEO, Kik