As the move to cloud platforms speeds up, the pressure is on to take advantage of bots powered by artificial intelligence (AI) – especially for IVRs.
Many businesses are at a standstill in adopting AI because they’ve done nothing to their IVRs for a decade or more. Their old IVRs are complex and slow to update, with mediocre customer experience, at best. But most are terrible. The State of IVR in 2018 noted that 83 percent of customers would avoid a company after a poor experience with an IVR.
I recently phoned my utility provider, and the IVR pushed me through eight different menu options. Each option took five to 20 seconds of listening time. By the time I got halfway through the eighth option, I had forgotten what the first one included – and I had to go back to the beginning. Consumers are frustrated by long IVR menu choices.
They’re even turning to online cheat sheets for ways to bypass a particular company’s IVR and get to a live agent.
Fear of change, even when it makes sense
Despite the evidence that customers are frustrated with IVRs, and the rapid decline of the old-school telephony, businesses are still reluctant to change. Some pushback occurs because of successful containment rates of IVRs. For others, it’s fear of changing menu options for customers who know exactly which number to press to self-serve.
One bank told me that they were reluctant to change because they have many customers who program their IVR options into their phones, including their PINs. Banks are exposing themselves – and their customers – to major security breaches, instead of doing anything about it.
While some try improvements like adding automatic speech recognition (ASR) with predefined expressions, they fail to recognise that it’s a short-sighted solution to a long-term problem. They need to fix their outdated design.
IVRs and the challenge of multiple intents
In traditional IVRs, customers select only one option at a time, and the IVR can process only that one intent.
However, most people multitask. Let’s say you dial into an IVR to change your address and open a new savings account. Then you remember that you need to add someone to your existing account. Typically, you’d complete one task and then return to the IVR or have an agent transfer you to another department to do so.
That’s because when those secondary intents come up within the conversation with an agent, the agent isn’t equipped to help. The secondary intent is often not dealt with, recorded or tracked. The customer still needs support, but the case is closed. And all that valuable customer information is lost – along with customer satisfaction.
Voicebots identify multiple intents upfront. They can handle many of them within the IVR and, if needed, pass all those intents on to an agent. Your IVR can become a conversational IVR, capturing context and vastly improving the Customer Experience through personalisation.
This is key to exceptional CX – and using Natural Language Understanding (NLU) within your current IVR makes it possible.
Voicebots and conversational IVR
Google led the modern revolution of conversational AI with NLU.
This technology makes it possible for a voicebot to hold a conversation and conduct back-and-forth questions, prompts and answers – without the customer having to use predefined expressions. In this way, every customer has a hyper-personalised experience.
Conversational IVRs go beyond understanding words as experienced with ASR, to determine what the customer wants and to help the agent understand and respond effectively. Machine learning capabilities enable these increasingly rich conversations – and continually optimise the IVR and improve the Customer Experience.
After the voicebot identifies the intents and self-serves where possible, customers can still go through a standard path within the IVR – or they can be routed to the relevant skilled resource to help them. Voicebots offer a massive opportunity to streamline the entire interaction process.
Let’s say I call my mobile carrier because I’m going on holiday and I want to know what the charges will be when I go overseas. With that one utterance of “I’m going overseas”, a voicebot would understand that this statement likely will require additional information.
The voicebot could ask: “Would you like to enable international roaming?”
If I answer yes, the voicebot could automatically process that request and then inform me of the expected tariffs. And, it can still pass this on to an agent if my questions are too complex. It’s a fluid, hyper-personalised conversation, and it doesn’t have to be complex.
You don’t have to change the entire IVR to use voicebots.
Voicebots move Customer Experience to the forefront
Voicebots not only solve long-standing IVR problems, they also take advantage of the data you already collect. Compare the advantages of conversational IVRs led by voicebots to traditional IVRs that put customer experience second to containment. The time savings, Customer Experience and overall improvement in operational efficiency blow traditional IVRs out of the water.
There is ever increasing interest in the role emotions play when managing Customer Experience in the contact centre.
At the same time, there is a drive to introduce technology such as chatbots to make customer service teams more efficient; removing repetitive tasks and providing ‘always on’ customer service. These potentially conflicting trends are happening at a time when the demand for customer service is growing, and organisations are fighting to differentiate themselves through their customer service offering.
A recently commissioned study by Forrester Consulting suggested that 90 percent of customer service leaders agree personalisation is core to the future of automation, and existing chatbot technology is stalling their efforts. The key challenge is to build simple yet personalised experiences for customers.
As Maya Angelou famously said: “People don’t always remember what you say or even what you do, but they always remember how you made them feel.” If your chatbot or AI solution leaves the customer feeling frustrated or angry because they have to put in more effort to get the answer to what they perceive is a routine query or task, all that is being achieved is an increased chance that the customer will look for an alternative supplier who can make this task easier.
In addition, quite often humans want to talk to humans. A study by PwC found that an average 74 percent of non-US consumers want more human interaction in the future and that 59 percent of all consumers feel companies have lost touch with the human element of Customer Experience.
Certainly, there have been strategies employed whereby chatbots are being disguised as humans which can only lead to frustration on behalf of the customer when they find they are being deceived and the bot cannot fulfil their needs for a more emotional or complex issue response. While customer views are constantly evolving, I still think Userlike got it right with their view on avoiding the ‘uncanny valley’.
Organisations need to be up front when a customer engages with them by disclosing that they are talking with a bot, and take advantage of the benefits that can be gained when effectively deploying it for more routine and simple tasks. In addition, they need to give the customer the opportunity to seamlessly switch to a human agent, without the need for the customer to repeat themselves. In short, make it easy, make it simple and, when the customer is speaking to an agent, make it personal.
No one can deny that AI is getting better and better, and chatbots will certainly have their place in our future. A well-designed customer-centric journey will allow the bots to tackle low level tasks, but companies also have to be cautious in blindly launching bots into the contact centre eco-system. When poorly executed the effect upon customers can be detrimental to their overall experience. It’s all too easy to deploy a chatbot that can get stuck in a loop, resulting not only in an increased cost to serve but also a decrease in overall customer satisfaction.
Hockenbury & Hockenbury in Describing Psychology (1997) described emotion as “a complex psychological state that involves three distinct components: a subjective experience, a physiological response, and a behavioural and expressive response”. Delivering customer service for an organisation dealing with often highly emotive subject of money, we have yet to find an AI solution that can effectively replicate the human touch our industry-leading customer service team can deliver. They can handle the simple routine tasks well, but then so can a well-designed FAQ or Help Centre. Until such a time as when chatbots can manage all three psychological states, there will always be a need for humans.
Human agents have a big advantage. They understand compassion, they can demonstrate empathy and they have their own shared experiences of everyday life which continues to become busier and more stressful for us all. In having this unique skill set, the human agent is here to stay and will own the complex matters where a human touch is needed.
The recent Genesys Xperience19 conference in Denver, Colorado, saw some of the most exciting developments in Customer Experience technology showcased to an eager global audience, and in case anyone was under any illusion about the future of CX – it involves AI.
The tech itself is dispassionate, and can appear benevolent to users as it cheerfully helps them along their customer journey. However, decades of pulp sci-fi dystopia has left AI with an image problem – no matter how helpful it may seem, some simply cannot shake the idea that bots might someday pull a Hal 9000 and see humans as inferior and deserving of subjugation…or worse!
Such fears ought to be dispersed when one discusses the details of AI technology with the real intelligence behind it – someone like Olivier Jouve, Executive Vice President of Genesys Purecloud, perhaps the planet’s most popular contact centre platform.
Olivier took on the role of PureCloud EVP in 2017, having spent over three decades honing his craft in pioneering customer sentiment technology development, including through senior positions at IBM.
His impressive resume also reveals a stint as an associate professor in computer science at Leonardo da Vinci University in Paris, and today, with a 150-strong AI team under him, Olivier is one, if not the world’s foremost authority on AI and its ability to make our journeys as customers easier.
He knows, in detail, how much his tech helps us in our lives – often without us realising it – but still the idea that AI will have a negative impact on humanity can cloud the vision and judgement of some sceptics who see it as an evil overlord-in-waiting.
Fresh from a timely Xperience19 breakout session on AI Ethics, Olivier took time out to chat with Customer Experience Magazine about his work, its reputation, and just how much AI is used for the betterment of our lives as consumers.
Speaking of the “creepiness factor” that some associate with AI, its access to personal data, and how it could be used in the wrong hands, he describes why being open and honest with customers, and letting them see the advantages with their own eyes, is the best way to go.
“We want the customer to know that we respect their data, and we need them to see what data we are using, so they are able to opt out if they so wish,” he explains.
“In the way we build our products, we do a lot of design thinking with customers to understand where the limit is. You know, what type of data they are comfortable with.
“And of course, you are being careful not to introduce any bias, which is something that’s very complex – not using any gender, or lifestyle, race…whatever, that could turn your model into something that is going to be targeting a specific minority.
“This wasn’t on the table 15 years ago when we were already scoring contact centres for next best action, cross-selling, up-selling, and so on, and using that data. Now there is much more sensitivity about how you use the data, and I think that’s actually a good thing, as it forces us to be clear from the get-go.”
Olivier highlights that those who are creeped out by an AI’s use of data, to the point where they will walk away from it, are a small minority compared to those who see the benefits and remain loyal to brands brandishing the tech.
“Companies which use AI the right way will enjoy great benefits, by being fair and respecting privacy,” he adds.
A common cause of ‘creepiness’ is the notion that a customer is unaware if they are interacting with a human or a bot on their journey with a brand, but as Olivier sagely states, that uncanny valley effect is being superseded by good old fashioned customer satisfaction when the AI does its job – and does it well.
“Me, personally, I don’t care if it’s a bot, as long as I get what I want, quickly, and with a great experience,” he continues.
“I don’t think customers care as much about the technology they use as much as the experience they have. I do think we should disclose that it’s not a human though – that should be part of the disclaimer. But at the same time, I don’t see that as something that should be discouraging people, who may think ‘oh no I’m not talking to a bot as I won’t get anything from it’.
“That’s also a danger of going to market too quickly with AI tech – some chatbots don’t provide the right experience. There are, however, things chatbots and voicebots can do very well, and I think we should double down on those.”
As Olivier points out, it’s not as if customers aren’t already used to interacting with bots on a less ‘intelligent’ level already.
“When someone asks for the balance of their bank account, they don’t care if it’s a human giving it to them, so people are already used to this sort of automation. We just have to be careful that if we go deeper with more complex things that users don’t get the feeling we are not responsive.
“People like empowerment, and chatbots can be great for that. But there are still some limitations, so we are not yet at the stage where AI is going to replace humans. We have chatbots that are very specialised and do things very well, but we need to find the right moment where you have to hand over to a human.”
Yet will there be a day when there is no human to hand over to? Will we fleshies be redundant in a future where all the work is being done by bots?
“I don’t think it will replace humans, as we are putting humans into something they are really good at, and so I see that more as a collaboration between AI and human – something we call blended AI. We can do sentiment analysis automatically, but it has limitations,” Olivier replies.
So what are these human skills that we can still feel superior to the bots on, and that customers still desire on their journeys? What’s the key difference that currently keeps humans in contact centre customer-facing roles?
“Empathy – we aren’t there yet,” says Olivier.
“Humans are great at it, and we need to think of the overall CX, the CX we want to provide. AI does things a human cannot do because you could not integrate all the different insights you have about a customer, but AI is really good at that.
“However, to take the conversation to the next step, at some point currently you have to hand over to a human. Of course, even humans need to be taught empathy in some cases!
“Perhaps in 10 to 15 years we will be able to train AI better in applying empathy, but for now that’s why we need this combination of human and technology.”
Ok, enough of what humans can do better than bots! It’s time to let Genesys genius shine, as Olivier outlines exactly why today – not years from now – AI is simply superior in most non-emotional ways to humans when it comes to steering customer journeys and earning the desired end result – superior Customer Experience.
“Our products work on finding the best agent for an incoming call or interaction, something we do very very well thanks to machine learning. The AI is understanding what the topic of the interaction is and uses historical data and a sophisticated decision tree to move the interaction forward.
“We want to optimise the customer journey, so we have a solution called predictive engagement. We can look at what a user is going to do on a website, view their navigation, and see at which moment he or she might need some help, and decide what the best outcome is for this customer.
“Once you have this interaction you can develop additional models which could be for retention or selling – more things we do very very well that a human couldn’t match.
“When you do next-best action and you have a customer calling who is ready to leave, our tech knows if you go for a specific action, there’s an 85 percent chance that this person is going to remain as a customer, for example.
“That’s from crunching a lot of data, gained through similar situations, and handling so much data – well that’s not something a human can do.”
Other than the aforementioned empathy, does Olivier feel his AI is lacking in any other areas that might be beneficial to overall CX?
“I think what AI does not do very well yet is go deeper into the conversation,” he tells me.
“We see breakthroughs when we look at things like Google Duplex, where, you can find yourself questioning if you are interacting with a real human or not, but an AI able to handle 100 percent of complex interactions? I don’t think we are there yet, though we are making a lot of progress.”
Despite my impatience as a customer to know when exactly this will be possible, Olivier sensibly refuses to give a date.
“I don’t like giving predictions, but I see how fast we are moving forward. I think Duplex was really a breakthrough – suddenly you see something and you think ‘wow’ – the voice, the type of interactions…it’s all very human.
“I don’t think it’s about developing the technology now – it’s about the right data and making it accessible. All of that is moving at an exponential speed. What’s really accelerating AI is that everything is in the cloud. Every single interaction from the employee or customer’s side – all that is feeding our AI platform. The more data we have, the more we are going to be able to power the customer journey.”
In the midst of such dazzling tech capabilities, it can easy to forget any regulation necessities to protect data and ensure AI is used for the good of humanity.
“It’s our responsibility to propose how we want to be regulated. It’s the right time to do the right thing,” adds Olivier
“Over 30 years I’ve been through a few AI ‘winters’, where AI had been at peak hype, but then died. I don’t want this one to die because some people are not responsible, so I will do whatever I can to make sure we are doing the right thing.
“There remains a fragility to this whole sphere, caused by the actions of Cambridge Analytica for example, which rocked people’s confidence in AI and data use, but I believe what we are developing at Genesys is promising, and beautiful, in a way that will not kill the hype this time around.”
So how do you stop your users getting fed up, and calling you instead? It’s all in your language. Find the right tone for your bot – and make it a master conversationalist – and you’ll save your customers from banging their heads against their screens.
1. Sound human (just not too human)
The Turing Test might once have been the standard we held robots to, but we don’t want our bots to pretend to be human any more. A Goldsmiths study found that almost half of Brits think it’s ‘creepy’ when a bot pretends to be real. We don’t want to feel duped – we want to know where we stand. In California, it’ll be illegal for a bot to pretend to be human as of July this year.
That doesn’t mean we want to have to interpret robot-speak, though. In a conversation with a bot, we want the same things we want in any conversation: empathy. Warmth. Signs that the other person is listening. One of the most natural ways we show empathy is by mirroring the language of the person we’re speaking to.
So writing more like you speak will serve you well here. Use conversational words, contractions, and the odd discourse marker like ‘ok’, ‘well’, and ‘right’ to help build a warmer relationship with your users (of course, there are as many ways to sound ‘human’ as there are, er, actual humans. It all depends on your brand’s personality, and the persona you’re after for your bot).
2. Set the parameters of your relationship
One of the biggest reasons your users are getting frustrated with your bot might well be because their expectations don’t match up with what you can deliver. So from your first interaction, be totally clear on what your bot can – and can’t – do. And pick the right points in the conversation to remind customers what those boundaries are.
Your intro message is the place to start. Nike does a nice job of this in theirs:
“I can help with an existing order or get you set up with the hottest shoes & gear.”
As does Western Union:
“Chat with us to send money, track transfers, check exchange rates/ fees, find agent locations, and more.”
Your bot not understanding a question is another good time to direct users to the right tasks: “Hmm, I can’t help with that. Here are a few things I can do…”
Set the parameters for the tone of your conversation, too. If you start on a subservient, apologetic note – “How may I be of assistance?” – you’re priming your user to have less respect for your bot (plus, if your bot’s female, using language like that could reinforce harmful gender stereotypes). Go for an adult-to-adult feel instead: “I’m here to help. What’s happening?”
3. Balance personality with getting on with the job at hand
Your users might come to your bot because they have a job that needs doing, but they’re more likely to stick around, come back, and tell other people about it if they have a good time in the process.
Take the time to craft your bot’s character: what’s their back story? What are their hobbies? What other robots would they be mates with in the playground? Turn that character into a set of guidelines that anyone writing for your bot can follow – like the ones we did at The Writer for Vodafone’s customer service chatbot TOBi.
Don’t let personality get in the way, though. No one wants your assistant to crack a joke at the expense of actually solving their problem. Save those snippets for the Easter eggs – hidden gems to reward you user with if they ask certain questions.
Ultimately, the smarter machines get at things like speech recognition and natural language processing, the less fed-up we’ll get as consumers. But all of that technological brilliance is wasted if your bot’s just rubbish at having a conversation, and without things like body language or visual cues to help guide that conversation, it’s your words that need to do all the heavy lifting. Pay more attention to those words, and your customers will feel more inclined to keep coming back – and less inclined to chuck their Alexa out the window.