← The Digital Marketing Podcast
The Digital Marketing Podcast
Message Me
00:35:38
The Future of Customer Service in the Era of Social Messaging and Artificial Intelligence
Loading summary
Transcript35 lines
- [00:00]
A
Welcome to the Digital Marketing Podcast brought to you by targetinternet.com hello and welcome back to the Digital Marketing Podcast. My name is Kieran Rogers and today we've got an interview which is all going to be about customer service and the fast moving pace online of customer service and all the tech that goes behind it and the gentleman we're going to be talking to. For anyone that's listening to back issues of the show, you may recognize him. His name is Joshua March and Joshua was on the show all about five years ago, kind of 150 or so episodes ago, talking about the tool that his company developed, Conversocial. But today, very lucky to have Josh back. He's written a book all about customer service in a 21st century and I'm very excited to have him on. Joshua, introduce yourself to the audience.
- [00:58]
B
Hi Kieran, how are you doing? Yep. I'm Joshua March, the founder and CEO of Conversocial, as you said, came on five years ago, which was really right at the beginning of our journey as a company. You first raised money and became a company in mid-2011. And we started the company really based on this vision that as social media continued to grow and become more and more important as a communication channel, it would have to kind of move away from just being managed as an island in the marketing scene, which at the time in 2011 it certainly was. And it would have to be deeply integrated into all of the relevant business units. And we were in particular really excited about customer service and we really believed that social would have to move into the contact center and be managed as a, as a customer service channel. And actually it was about, it was kind of around that time that we had started, we'd been hired by Tesco who really shared that vision. They were one of the first kind of early innovators in the kind of customer social care. Social customer care space and hired us to help them move social into the contact center. So I started personally traveling around into these contact centers and watching the difficulties that customer service agents were having trying to resolve complex cases over multiple messages with very basic tools that didn't really have proper workflow. And I saw the challenges that the managers and supervisors were having trouble trying to understand agent productivity and cost resolution, all this kind of stuff for software that just wasn't really focused on likes and clicks and shares and kind of got a really good understanding of just the sheer complexity and difficulty of building a kind of enterprise grade contact center platform and got really excited by that. So we between the kind of rapidly shifting worlds of social media and now mobile messaging and the needs of the large enterprise contact center. And that's very much kind of what we've done five years on. Obviously there's lots happening now and there's more shifts happening, but Here we have 250 customers today tend to be kind of large consumer brands who sell direct. So a lot of retail. Tesco still a client, Waitrose, Sainsbury's, M and S, all the big retailers in the uk, similar in the us so people like Macy's, Container Store, Toys, Toys R Us and also airlines. And travel is a really, really big vertical. So like Ryanair, Aer Lingus, Iceland Air, Alaska, Alaska Air. In the US people like Hyatt Hotels, Hertz Rental Car. So tends to be on the larger side because they're the kind of brands where the volume of customer service they're getting through social media means that they need pretty significant teams just dedicated to it. So, you know, I like some of our biggest customers have got 150, 200 agents who are full time just responding to service issues through social media.
- [04:06]
A
Now from my perspective, conversational really helped customers deal with that first big shift where consumer conversations were really en masse moving online. It did happen quite suddenly. I remember at the time I was working in a skincare company, we had a big call center and we, we were set up to deal with customers phoning in and speaking to us about all their customer service issues. And everything moved online into all these multiple social channels. And actually it was a real headache at the time. There wasn't a lot of tools around like there are now to really help you manage these multiple conversations in multiple social channels. And conversational was a real game changer for us. It actually shifted the focus of social media or away from the PR team where it had been, into the, you know, the heart of customer service. Now you've written a book about the customer service in the current landscape. I've spent all day reading it actually. Brilliant book. I couldn't put it down. And the book predicts the next big shift in customer conversations. Can you share with us what you think this next big shift is going to be?
- [05:22]
B
Sure. So if we go back a little bit and talk about that initial shift that you just mentioned, social media kind of customer service on social media really started as a kind of escalation channel. So someone had probably had a bad service experience or had struggled with getting a good experience over the phone or in store or whatever it was. And they kind of turned to social public Social media as a way of escalating it up and getting more attention to their issue. And as people found that to be really effective, often they started just turning to that as their kind of primary way to get help. So next time they had an issue, they go, well, these guys were terrible over the phone, but they responded really quickly and resolved my issue over Twitter. So now, instead of phoning at all, they just went straight to Twitter or straight to Facebook. And, you know, as those platforms increased the functionality around their private messaging, often the brand started encouraging people when they had an issue to DM us or message us and increase. And we started to shift over the kind of 2016 where the volumes of private just kind of kept going up and up and up. And actually for some of our clients, kind of taking over public, there's a lot of their customers who've been using them. Customer service knew that they'd get a quick response and it was super convenient and easy for them just to kind of message them or DM them. And so it became less of a kind of forum for public complaints and just as a way of getting a direct access on your phone to the customer service team. And the kind of rise of that private messaging was really interesting for us and really interesting for brands because a lot of brands have been scared of promoting these channels, social care, right? Because they didn't want to encourage public complaints. But as it became easier and easier to promote private messaging, that kind of complaint was removed. And a lot of brands started to recognize, oh, you know what, these channels are actually better for the customers. They're more convenient, much easier for them to reach and access and use on their phone. They have these continuous conversations, so you never have to kind of repeat the same thing for different people. There are lots of reasons which I delve into as to why they're so good. And so companies started over the last year or so actually promoting these private messaging channels as a cat channel. And the stats on that are pretty impressive. We had one retailer, a big retailer in the uk, who I can't mention, they put these links on their Contact Us pages in August, just before the bank holiday weekend. And before that point, social media in general had made up around 6% of their inbound customer care volume. Once they put the actual links on their Contact Us pages, within two weeks, their email volume had dropped by half and their messenger volume had doubled. And over the bank holiday weekend, 21% of all of their inbound service volume was coming through social media. Right. So it's clear that like Customers wanted to use that and customers preferred it. It's just that there was a lot of customers who didn't even realize that that was a channel they could use. And I think that's this kind of rise of private messaging as a just kind of primary care channel that companies are promoting is something that we're starting to really see come into force now. So I think we're going to see a really big shift where more and more companies are promoting messaging, and it starts to really overtake certainly all of the other digital channels as a. As a primary care channel.
- [09:10]
A
For years, people in customer service have been focused on delivering, you know, absolute over the top customer delight. That is, if the customers are actually able to get through the phone queues to actually speak to them. Now, your book argues for something actually far more simpler. This was one of the points that really struck me about the book because you sort of. You're really slaying a sacred cow there. It's like, no, wow, that's amazing. But actually, you know, you present a good argument. Can you explain what this something simpler is and why you think it's so key to this revolution that you see is taking place?
- [09:44]
B
Yeah, absolutely. So, I mean, I had a similar experience a couple of years ago when I read Effortless Experience by ceb, who now owned by Gartner. And it was a really, really impactful book for me. And I started kind of buying it for everyone I knew and buying it for clients and this kind of stuff. And in it, essentially, there's a really deep dive into why people go to customer service, the impact on things like NPS after they have service interactions, why they phone versus use online, and stuff like that. There are a number of key findings. One major one at the heart of it was that almost every service interaction that someone has results in them decreasing mps. Almost every service interaction. And the things that cause that drop in MPS are often just the very, very simple things that increase effort and increase work by customers.
- [10:52]
A
So for those that don't know, Joshua, what is nps?
- [10:56]
B
Our NPS is Net Promoter Score. So it's a very commonly used metric. I think it was first created by Bain. A lot of businesses use it as their kind of core metric for how happy their customers are. It asks a question of saying, like, how likely are you on a scale of like 0 to 10 to recommend us to your friends? And then it has a particular methodology where it basically looks at who are the people who are right at the top, 8, 9, 10 who are kind of promoters compared to I think who are the detractors? Which is anyone kind of, I think 1 to 5 and 6 to 7 is neutral. And it kind of looks at the difference between the promoters versus the detractors to create a score. So that score can be from like minus 100 to plus 100. And it's often generally in the contact center world, companies will either use MPFS as a key metric for like after the service interaction, is their NPS higher or lower? And that's a kind of good measure of how that service interaction was. And some other companies use just customer satisfaction, CSAT seller specifically how was this service interaction between 1 to 5. But there are a lot of businesses that actually their entire business will be bonus on what their overall NPS score is. And all the things they're trying to do are increase mps. But the kind of trap that a lot of companies go into in customer service, and this is what comes down to the effort stuff, is that they have this kind of conflict where they go, all right, well, when someone gets in touch with customer service, we need to really do whatever it takes to make them happy because we care about mps so much. But at the same time, phone calls are so expensive and they have this really big pressure to decrease costs. And often the way that they do that is they go, okay, well we'll hide the phone number, we'll make it harder, we'll make it harder for people to like reach. We're going to try and decrease phone calls by putting a complex IVR with seven different levels before they can get to someone. And actually to get the phone number, they're going to have to go through like way through Pages of Health FAQs first. But if we do get through to someone, we'll try and go over the top and do the best experience. And what the data actually shows is that if someone has a problem or a service experience, they're not happy already. Right. And if you go really, really over the top of helping them and go creating this incredible experience, you could have a slight improvement on their NPS score. A slight improvement. But if you've made it hard work to get to that point or put any barriers, if you're having to repeat the same thing to different agents, for example, or they tried to get the answer online, struggled and ended up phoning because they jumping or anything that required them to jump through hoops and put hassle in front of them has a four times bigger impact in a negative direction. And so actually most companies are like measuring, you're trying to do the Wrong thing. And actually they're lowering MPS with every service interaction. And rather than trying to deliver this kind of over the top delight when someone finally gets through to you, actually if you just focused on trying to make it as easy and effortless for people to get their answer as quickly as possible, as easily as possible, without jumping through any hoops, without having to do all this kind of hassle, then has a much more positive impact.
- [14:32]
A
There's a really good phrase in your book which is just hit the easy button, which I thought was great. It's just the thing, I think that all customers wish businesses would provide for them. But actually a lot of businesses, particularly online businesses, actually can be really terrible at doing this. You know, it can be, can be really hard to actually get a human response or to know where to go if you want to get some help from an actual human or get advice when dealing with many of these.com mass market tools that are out there.
- [15:06]
B
Yeah, yeah, 100%. And I think especially in today, right, I mean, effortless experience was a few years ago. And since then this has just become even more important because we just live in this like app world now, right, where we're like pressing a button on our phone and suddenly a car arrives and takes us anywhere we want. We'll press another button and like a hot meal arrives at our door. You want to go on a date, you just open up Tinder and you like press a few buttons and you have a date. Like almost anything in this world now is like accessible with a few clicks on your phone. Apart from if you have like a service question, at which point you're like wading through reams of stuff on a website, you know, waiting on hold for 30 minutes. It's just like so painful. And a lot of that is because I think companies have just been focusing on the wrong things and I think that kind of the focus on effort. And when companies kind of switch their focus to effort and the lens of how they think about service to effort, then that really starts to show why channels like mobile messaging are so important, right? Because if you can just whip out your phone and message a company, that's super easy for a consumer. And because it's a continuous thread, right, there's no need to repeat what you ever said, what you say to anyone. It's not like live chat on a website where you have to be sitting there on your computer. It's like texting a friend. You can message, put your phone in your pocket, come back to it later. The agent could follow Up a week later and it just pops as a notification super, super easy. And the data is there now, which shows that it's that ease of use which is actually going to have the biggest impact on nps.
- [16:49]
A
Yeah, I agree. And there's a really compelling argument for that in the book, which you've backed up with some really great case studies. With my experience working in lots of businesses, you do realize that actually phone and email really were not built for effective many to one mass communications. It just doesn't work. Every place I've worked, customer service centers get so easily behind on emails and constantly struggle to deal with call volume at peak time. I guess it's because, you know, when problems occur, it happens in fits and starts. It's so difficult to scale up and down quickly enough. It just makes you realize actually it's almost intrinsically broken. You know, increasingly we don't even use our, the phone function of our smartphone when it comes to actually contacting businesses.
- [17:41]
B
Yeah, yeah. Well, it just takes so much time and you have to be fully committed to only that for a long period of time. And especially if you're just trying to fix something that was broken in the first place. No one wants to commit an hour of time to wait on hold and speaking to agents about that.
- [17:58]
A
No, indeed. Or to wait 24 or 48 hours for an email response, which in some circles would be considered quite quick.
- [18:06]
B
I know, yeah, it's crazy. It's really crazy.
- [18:08]
A
Do you think the writing's on the wall for phone and email as a customer service medium? Is there any hope for these beloved devices that have graced our desks for so long?
- [18:19]
B
Well, I think the devices will be, but definitely the medium, the channels will be different. I mean one of the data points I mention in the book was a study done by Ofcom in the UK where they looked at mobile phone usage and it was 12, 15 year olds at the time who are now probably more like 16, 17 versus kind of adults I think were like 18 to 34. And if you look at adults today, adults were like phone and email was probably like 50 or 60% of what they did on a smartphone and the rest was kind of social media and mobile messaging and photo apps. When you look to teenagers, phone and email were like less than 5% combined. It was like crazy, just like disappeared into nothing. And almost the entire use of the phone in front of me was mobile messaging, social media, photo apps. Like that was just like that's how people communicate and interact with each other.
- [19:19]
A
Those Ofcom results really Fascinated me. You know, when you compared it, the 12 to 15 year olds are using phone and email about as much as current adults use photos messaging. And that's, that's shocking. I could relate to that. As you know, I, I guess as I'm a little bit older and longer in the tooth, I hardly ever use photo messaging. But to imagine a world where people used phone and email about as much as I use photo messaging, you know, that, that was a real eye opener. You start to think like, wow, that's, that actually predicts a massive shift in, in customer current behavior. And yes, okay, they're 12 to 15 year olds, but if that's what the current norm is for that generation growing up, that's going to translate into some huge shifts later down the line.
- [20:12]
B
Yeah, yeah, completely. But if you look at a contact center today, they still receive letters. Not a lot, but they all still receive letters. And I wouldn't be surprised if a contact center in the Future is like 80, 90% messaging. But they're still going to get the occasional phone call, they'll still get the occasional email, they'll probably still get one or two letters. There'll always be some holdouts. These things aren't going to disappear completely, but I think they are going to be dwarfed in terms of volume.
- [20:47]
A
So one of the challenges in this space has always been that diverse number of channels that consumers want to interact with businesses through. Is it one of the big challenges ahead with so many chat clients and customers still wanting to communicate through email and through the phone? Is there ever going to be one ring to rule them all? And will current tech really support that?
- [21:10]
B
Yeah, I mean the nice thing about messaging is that it tends to all have pretty much the same kind of form factor. So, you know, certainly when it came to public social media, right. The way that public Twitter works is very different to the way that Facebook pages work. Right. And it's very different to the way that comments on Instagram work or YouTube work. Yeah, all this kind of stuff. And so that's kind of a bit of a pain. But when it comes to the private side, they're pretty much all identical, right. It's all this one to one messaging that can combine real time and asynchronous, often building kind of bot automation platforms in the same way that we'll come on to. And there's that and there's also a push to kind of try and move web chat to be asynchronous and more like messaging. Right. So I think we're going to get to the future where even with proliferation of messaging platforms, they can all be managed in the same way from, from a kind of customer service perspective, you know, so you can have the same agents handling issue coming in from all of these different messaging channels fully mixed together and blended, because you don't have to worry about all these different kind of public form factors. And is it a retweet or a hashtag or you have a comment or reply and all this kind of stuff which gets so confusing on the public side.
- [22:41]
A
So how does artificial intelligence work into this mix? And is it more machine learning labeled up as artificial intelligence?
- [22:50]
B
Those things are kind of the same. So yeah, I think where artificial intelligence is kind of a slightly nebulous term, I think one of the best lines I heard about it is that people use artificial intelligence to describe things that haven't yet been implemented. But as soon as something is actually implemented and used, everyone calls it machine learning. But I think where it does need to be separated from is the kind of bot platforms because there's all this hype around bots over the last year, year and a half. And pretty much all of the bots that have been created use the bot platforms within the messaging apps to create very simple rule based bots that aren't really any different to kind of chat bots that have been around for decades. The bot platforms in the messaging apps just make it easier to build them, much simpler to spin them up. Let me get into that a little bit because I actually think that AI is going to be hugely, hugely disruptive to the customer service industry over the next five years. And I think that messaging is going to play a core role in that.
- [24:11]
A
Okay, so just explain how you think that's all going to play out.
- [24:15]
B
Sure. So firstly, today there isn't really much AI used in customer service at all. We use elements of machine learning for routing and optimization and stuff like that. But in terms of directly responding to the customers, there's very, very little across the industry. I saw a stat recently where in the live chat world, something like less than 5% of companies use chatbots. In the live chat world, there's two reasons for that. One is this thing that historically people have just been, you know, building these simple rule based bots that really don't work very well for an unstructured conversation. They just break down. And the second issue is that in the live chat world, you know, you're in a real time conversation with this bot, you're going to, you just have to stay with the bot. Until the bot breaks at some point, and then there'll be a kind of painful handover to a human. But in the messaging world, the asynchronous nature means that you don't have to have a fully standalone chatbot. Instead you can just choose to automate a response on a message by message basis. So some messages a simple question, and you can have a system that recognizes that question, recognizes that it's common and knows the answer to it, and is completely certain that it can answer. And it could just answer it instantly without that ever going to an agent. If the person then comes back with a slightly more complicated follow up, then you might have a system that goes, oh, I think I know the answer to this. It could be one of these two. And because it's not a fully real time channel, we can just pop it over to an agent's queue. Agent can spend 30 seconds looking at it and going, oh yeah, this is the correct answer. And then that gets pushed down. And if the customer then comes back upset about whatever that was and is kind of emotionally distraught, the system can look at that and say, I'm sending this straight to a human to deal with. And an agent can look at it and spend a few minutes crafting a real response. And after that, the customer comes back again with something simple. Potentially it's automated again. And because of that, it means that you can start implementing automation and AI solutions that only have to automate a small percentage of the conversation to be able to have a really positive impact. And that's positive in terms of getting a faster response and resolution to the end customer. Because part of it is automated, increasing the efficiency of agents, because suddenly they're not having to waste time answering very simple questions. They can focus on the more complex or emotional questions. And it never has a negative impact on the customer because they're never getting frustrated, because if the system doesn't understand them, they're never going to have a situation where it's like, sorry, I don't understand you. Instead, behind the scenes, it would have just gone to a human.
- [27:22]
A
So within this world, and you'll have spent a lot more time looking into this kind of data. Is there anything similar to Pareto's 8020 law? With 80% of the work being generated by around 20% of the questions, does it stack up at all like that?
- [27:39]
B
Yeah, I mean, it definitely depends on the individual company, on the complexity of the issues and everything else. But certainly there is going to be that kind of power law curve where there's going to be a lot of questions which are quite simple and which can be automated. And then there'll be a kind of, you know, increase dwindling number of more and more complex questions that will are going to be harder and harder to automate. We a company that I'm friends with here in New York is called Work Fusion. And they, they're a deep learning AI company that essentially apply this exact same method but to kind of behind the scenes processing and business processing work. And what they find is that they'll go in and on day one they'll just implement some simple kind of rule based automation and that by itself will say kind of 15 to 25%. And we found similar. So we do, we build what we call a kind of visual IVR system where we'll kind of automate the first couple of messages when someone comes in to a private messaging thread, ask them what kind of issue they're having. Based on what they say, we'll be able to then route it to the right agent and we'll also automate the next message saying for that type of issue you need to give us your order number or something else. And so it only goes to the agent when all that information is collected. And again that just saves up to 15% of that simple back and forth at the beginning. And then once they start implementing deep learning that is actually learning based on historical actions and behavior by the agents, then over the next few months workfusion find that that will replace another 20 to 30% so that starting to replace up to like half of all of the work. And they have seen case studies of up to 80% automation over like a period of time and constant training. And people are doing that in this kind of behind the scenes processing, kind of robotic processing. We think it's really exciting to start implementing same kind of thing in customer service and messaging in particular makes it possible in a way that no other service channel has enabled in the past.
- [29:54]
A
You make a really good case for that in the book. And it does seem to be in tune with the fact that we sort of have two technologies really coming to fruition at the right time. I guess there will be some detractors there who say well, what if I don't have a smartphone? Or what if I don't want to interact through Facebook, messenger or Twitter? What would you say to that?
- [30:15]
B
Yeah, I mean there are always going to be some of those people. What we actually do for some of our clients is support messaging over sms which has been used for a number of years as a kind of one way customer service channel, giving notifications, flight updates. But not many people realize that it can be used as a two way customer service channel. And SMS is pretty basic, it's just text, it doesn't have the kind of interactive elements or UI of the messaging platforms. But it does work in exactly the same way from a kind of workflow perspective. And so again, you can have a single agent who could be handling stuff on messenger, on Twitter, DM, on WhatsApp and on SMS and can manage the workflow in an identical way. And that means that it's a really good channel to promote out. Again, if you're in that situation where you think there's a decent amount of customers who you'd like to support through messaging, but who don't have smartphones or in other situations may not have data. So like we have some trained companies in the UK who do SMS customer service through us and they do that because, you know, the trains are going through pockets of places where they don't have data but they want people to be able to contact them.
- [31:40]
A
One of the things you champion in the book, which I thought was brilliant, was getting a lot more people trained up so that when it gets busy, you know, you can really draft in kind of a territorial army of agents to help deal with customers problems from all around the business and respond at scale in a crisis. I think lots of organizations could really benefit from from that. Particularly if you only have maybe five to 15 people in your customer service team and you suddenly get hit with, I don't know, two, three thousand or maybe twenty, thirty thousand messages now. How do you cope with it? Well, you plan for it.
- [32:17]
B
Exactly. Yeah. It's about training the agents and having those plans and processes in place and.
- [32:22]
A
It'S so much fun to do. Who wouldn't want to learn this stuff? I think speaking to customers is brilliant. More people at every level of business should do it because custom that the whole point. So which brands do you think have really got this now? Do you know who are the brands that we could look to for sort of best in class experience in this space?
- [32:45]
B
Yeah, I mean there are lots of. What we tend to find is that there's no one brand that's doing everything perfectly right. But there are a lot of the brands which have really been pushing forward in this and innovators and doing a lot of exciting stuff in the space. Tesco continue to be pushing the envelope on this. Five years on they invest really heavily into social care and mobile messaging as a care channel and see it as a really important way to connect with their customers and obviously get quite a lot of kudos for that as a result. Hyatt Hotels are another one who just kept pushing the envelope from the early days of promoted messaging on their website in their mobile app, built integrations into their CRM system so that when you message them they know which hotel you're staying in, which is a pretty awesome experience if you've connected your account as a loyalty member. There's a lot of airlines who are pushing hard on this. Valeris is a Mexican airline who super innovative, growing really quickly and experimenting with using messaging as channel not just for service but for sales. Experimenting with different bot technologies, Alaska Airlines and Icelandair as well. Also investing super heavily and doing some really cool stuff. So lots of pockets there. Airlines in particular, we tend to see a really forward thinking in their approach to social, so definitely worth checking out, but lots of retailers and hotel companies doing exciting stuff as well.
- [34:18]
A
Well Joshua, it's been really great talking to you and thank you so much for really giving me the opportunity to read your book. Actually I don't normally read books on business from COVID to cover, but this one really grabbed me and I would highly recommend it now. It's called Message Me and it's by you, Joshua March. If I want to find out more about the book or ways to find out things more about conversational and what you're up to. How do I do that? Where do I go?
- [34:47]
B
Absolutely, yep. You can go to my own personal website, joshuamarch.com or follow me on Twitter just oshuamarch and that's M A R C H like the month and of course the conversocial@conversocial.com.
- [35:07]
A
Thanks for listening to the Digital Marketing Podcast brought to you by Target Internet. If you're investing in your digital marketing skills, take a look at our free Benchmark skills test and look at the wealth of online learning we provide to help marketers get up to speed and stay up to date. Just visit targetinternet.combenchmark.