The Post-Covid CX Powered by AI
Julie Lawson: Thanks for joining us for this amazing what the web episode, everyone, The Post-COVID CX Powered by AI, brought to you by Customer Experience Update and sponsored by SmartAction. My name is Julie Lawson, the webinar coordinator for this episode, and I’m excited to bring you some Gartner insights about customer experience. I’m really looking forward to interview and talking with our speakers. If you miss any of today’s episode, have no fear because this will be recorded. You can access the recording by clicking the link in the chat box, which will show up around the end of the show.
On that note, please feel free to ask questions during today’s session. Engagement from you guys is the key to a great webinar after all. You can do so by submitting them into the question panel on the right side of your screen. Aarde, Brian, and I will go through them throughout the entire presentation, and we’ll have time for Q&A section at the end. My lovely colleague, Erin will be fielding your questions today. She’ll be happy to answer any questions you might have. So pull up the question panel and say hello to let her know that you’re listening.
Lastly, if you’re having any audio issues during today’s presentation, you may choose to dial in by phone. You can do so by selecting the more button in the right upper portion of your screen and then select the switch to phone option. So today I am talking with Brian Morin and Aarde Cosseboom. Brian is the head of marketing at SmartAction where their obsession with creating frictionless customer experiences with conversational AI has led to its standing as the top-rated virtual customer assistance solution on Gartner, Peer Insight, and Distinction as the leader in AI, enhanced self-service by Frost & Sullivan. SmartAction operates AI-powered virtual agents for a hundred plus brands, including AAA, Office Depot, Hyundai, ADT, DSW, Choice Hotels, Penske, and more.
For the last decade, Aarde’s professional experience has been focused around contact center technology, servicing businesses, and consumers through cloud-based SaaS products. He enjoys working with large global customer service teams to increase and leverage the following areas of business, leadership, customer experience, customer success, operations, technology, customer service, recruiting, and people development. He has won multiple awards in the cloud technology field that includes best use of cloud technology and best voice of the customer product. So without further ado, take it away, guys.
Brian Morin: All right. Well, thanks, really quite an introduction. Of course, happy to have everybody here who’s joining us for the conversation today. We’ll be doing this a little bit more podcast style, I should say between Aarde and myself. So of course, we have set content. We don’t know if we’re going to get to it all. We’ll see how far we can get. But we like this to be audience driven as much as possible. So as we go along here, just chime in with your own question or comment as we go along. All of us are living right now in what feels to us is a very different world. So we’re speaking specifically about this kind of emerging narrative that we’ve begun to see asking about a post-COVID CX. We’re taking a very direct slant on that, on the AI-powered CX post-COVID,
Here, just a minute, we’ll just address first, is that even a thing? Let’s look at some of the actual real impacts that have been happening that have driven that narrative, and let’s get into what those are around business continuity, strategies that were not necessarily as much of a priority before, recession proofing strategies that we’ve begun to see and emerge.
So we’re taking a very specific slant around the AI-powered CX on this for a couple of reasons. One is you should see here just at least on this slide, if you don’t know, you’re not aware of SmartAction or know who we are, this just at least will give you the context. We design, build, and operate AI-powered virtual agents for more than a hundred brands that ranges from midsize companies to the Fortune 500. So it means that we’ve done a lot of end-to-end customer journeys from press one to, how can I help you today?
That’s a journey that we did with Aarde and TechStyle, who’s going to be sharing because we’re coming from the point of view, having operated in the AI-powered CX for more than a hundred brands, so we can come to the table and talk about what types of impacts we’ve seen COVID had on that and the trends that are emerging from it, and Aarde, of course, on the front line of this, within his own company, TechStyle can communicate that.
So for us you can imagine, you’re operating the AI-powered CX for some of the most trusted brands in the world that we’ve experienced. Every pitfall and peak are manageable over the few years. We do omnichannel, self-service. We deliver the full conversational AI technology stack and all the end-to-end CX services around that. So organizations can outsource to a partner, as opposed to trying to build up their own team and do this themselves. But the focus of this webinar is on voice because, at least from this standpoint, it’s where 80% of our customers start on the voice side. It’s where the see the biggest, most-immediate ROI. It’s far and away the most complex channel to automate well.
Once you have, it makes scaling any application against any interaction digitally to chatter SMS pretty straightforward. So Aarde, let’s talk just a little bit, if you can, just about yourself and your own street cred to your colleagues who are listening in, and then we’ll jump into a little bit of this conversation.
Aarde Cosseboom: Yeah. Absolutely. Thanks, Brian. Thank you to everyone on the SmartAction side for putting together all this content. Thank you for Brian and team and everyone who’s hosting this webinar. Julie gave me a great introduction, so I won’t go into too much detail there. A big thank you to all the attendees too. I know that in this post-COVID, we’re all kind of hitting our limits with webinars and Zoom and virtual meetings. Thank you all for taking the time to either watch this live or listen to the recording afterwards.
Of course, the conversation never really stops in this webinar. It keeps going and going. So reach out to either me directly or to Brian or anyone on the SmartAction team to ask for questions. SmartAction has done a really good job on their website to put together use cases and case studies.
I’m just one of very many of those use cases that they brought on for this webinar. So a little bit of background of who I am, other than the areas of focus, which you already heard and how long I’ve been in this industry. If I were to wear a NASCAR jacket, these would be the logos that I would put on them. These are the current companies that I’m either working for or currently engaged with help with their self-service. If you want to go to the next slide.
Just a little bit of a note, I recently wrote and published a book available on Amazon, and it’s sectioned into three different categories, but the first category that I talk about is enabling better service through AI, self-service, and automation. So talking not only about technology, but a vendor and partner relationships and how to create an ecosystem which delivers hopefully a better service to your customers on the end front. Then on the next two slides.
Brian Morin: Yeah. Well, Aarde, please. Yeah. Tell us a little bit about TechStyle.
Aarde Cosseboom: Yeah. On the next two slides, this is a little bit about TechStyle and our customer journey. This is our case story. I currently work there. I oversee the technology side for customer service and customer success. We have five different brands. We are an online e-commerce retail company. We also do memberships. We are very heavy on the membership model. We partner with celebrities to do the sales and marketing through social media and also through soft media. Some of the brands you may have heard of are Fabletics, FabKids, ShoeDazzle, JustFab, or Savage X Fenty.
Then zooming into that a little bit more. This is the amount of volume that goes through our contact center on a regular basis. We’ve got 5 million members globally, 12 different countries, seven different languages. The majority of our volume is over the phone channel. We have 6 million phone calls per year. So comparing that to the number of members we have, that’s a little bit more than one phone call per year, and we have 3 million chats per year from our members. So really, we have two touch points with our members annually, and we want to make sure that those touch points are extremely high level of success and high level of customer experience.
So choosing a self-service vendor is extremely important to us because it may be the one or two times that we talked to them in a 12-month period, and we want to make sure that they have a great experience as they go through that.
Brian Morin: Sure. Thanks, Aarde. Now, we’ve all been hearing this chatter, the new post-COVID CX. As I was first seeing that and hearing that, the skeptic in me asked, “Well, first of all, is this just a trending narrative, or is it something that’s actually real?” If it is something that’s actually real, then what is it? There are a couple CX trends that we’ve identified as new that I’ll address on the next side and banter that with Aarde. But for the most part, AI adoption trends, the self-service trends, they really have been going on for a while. There’s nothing new about that.
But what is new is how COVID seems to have accelerated the pace of change in some of these areas, particularly over-the-voice channel. The first day I speak to is that there’s a lot of, hey, let’s build it in digital first, where we can play in a sandbox, and if we get it wrong, it’s really just not going to harm our CX, and we can iterate and try. Some of the organizations have been going, but then suddenly in the pandemic, oh crap, our biggest problem is in voice, and we need a solution for that right now. Aarde, what’s your take on this?
Aarde Cosseboom: Yeah. As you can see, the majority of our volume is over-the-voice channel. It is the traditional channel of choice for our customers. We are a B2C. So business-to-consumer or business-to-customer direct. Customers really like to just pick up the phone and resolve something quickly. So that’s the channel of choice. In this post-COVID environment, where everyone’s quarantined at home, they’re looking to either multitask and solve some of their personal issues with companies much like ours at TechStyle. But they’re also looking for quick and short self-service options over multiple different channels, but specifically over the phone channel because they don’t have the time to really focus in on those problems because they’re probably juggling things like work and childcare and other things that may be distracting them while they’re working from home.
Brian Morin: So Aarde, we’re going to drill more specifically the AI-powered CX around this. We were having this conversation, I can’t remember last week or the week before. We thought, “Hey, this is something we really should be turning into what we’re doing now.” I just asked you just for candidly, you’re on the front lines. You’ve been on the front lines for quite some time. You were an early adopter. So when you look across the landscape at colleagues, that might be falling behind, or we’re getting, I should say maybe long in the tooth, if their approach or strategy. Are there certain areas that are most common to you that stand out?
Aarde Cosseboom: Yeah. We’ve had AI-powered chat bots and voice bots since 2015. So we were an early adopter, bleeding edge right back, when you go to conferences, even before everyone, the buzz was AI, chat bots, voice bots, all these automation, self-service agents. But it’s been there. It’s been lingering there for quite a while, and we’ve got a lot of small niche players that think that they’ve created the best solution, and it may be a good stepping stone or maybe the best solution four years ago. But as AI has evolved and as customer experience has changed or at least customer expectations around experience has changed, not a lot of the tools have changed to match that.
So to answer your question, I think that’s where most companies fall a little bit short. They have this thing on the roadmap that says we need to deploy a self-service tool. The first thing they look towards, maybe the easier route, which is a chat bot, which isn’t their primary channel. Maybe that’s only 20% of their volume or 30% of their volume, where really phone is their volume. They go through this process. They deploy something. Within a year, within six months, it’s already outdated. But they checked off that box, and it’s been sitting there for a year for four years, and now it’s become the stale product that’s just been sitting there, and there’s a little bit of a hoorah when it was launched, but now it’s, how do they rip and replace and make that a better experience, and how do they deploy it across channels that are more complex and have higher volume.
Brian Morin: Right. So if you were living in this new COVID world or post-COVID world, what does that new world look like? I think to answer that, you have to examine, well, then what has been the real impact that COVID has had on the ad powered CX or at least our awareness of it? The first one that I can speak to simply just because we do manage the AI-powered CX for as many brands as we do, the very first thing that we put at the top of list and not all of our customers because none of our customers were hit the same way during this time, but for a lot of our customers, there was a crazy spike across certain interactions for some of our customers that either didn’t even exist before, or they were only low volume.
So now, we’ll change that contact center. Leaders have to play almost this game and this what if scenario to make sure that they’re planned against what they may not right now be ready for. So a good example here for us is we went live just a few months ago with Choice Hotels. They’re the second largest hotel in the world. They are the umbrella for almost every budget chain you can think of. We handle all their reservations. We handle some of their outlier call types that don’t get much volume.
But when the pandemic hit, some of those outlier interactions increased by 400%. In fact, I did a webinar with them last month, and they had mentioned that they hadn’t… If they didn’t train the virtual agent for those interactions prior to the pandemic, they would have lost a tremendous amount of revenue because they were suffering Asian availability and long hold times at the very moment that those call spikes were happening, and that would have put additional pressure on their contact center that they would not have been able to handle.
Aarde, I know that since you’re not a brick-and-mortar retailer, maybe you didn’t see the same type of unexpected call types, spikes, or interactions in your business.
Aarde Cosseboom: Yeah. A good example for this. So we’re in fashion and e-commerce. So we don’t have brick and mortar. But we do have multiple different brands that we’re affected by COVID. The brand, Fabletics saw increase in sales immediately because everyone went from going into the office to work from home. So, everyone needed comfortable clothes –
Brian Morin: Aarde, I haven’t even changed out of a pair of joggers in the last four months. I hear you.
Aarde Cosseboom: I’m definitely in shorts at the moment. That that brand went like crazy. So the problem we had with that brand was an increase in volume. How do you serve without self-service in place, without bots in place? How do you serve customers in the way that they’re traditionally expecting a response without having the ability to forecast headcount and hire tons of people for those unexpected call spikes interactions.
Then the other way around, we have brands that are business casual apparel brands. So JustFab and ShoeDazzle, high heels and business attire, which the volume went down with those brands, and we thought, “Oh well, the volume in sales went down. So the calls will go down.” Actually, the calls and the emails continue to flow in because since we’re on a membership model, everyone wanted to call in and pause the billing for the next month or two months or three months. So we were in the call center with increased volume on both sides and both business units. One was going up and one was going down at that time.
So just something to think about, and then as we progress out of COVID in this post-COVID world, if you did have a loss in business and loss in contact volume due to the pandemic, how are you going to hire staff backup without self-service and technology in place to handle those unexpected calls when the pandemic is considered to be over?
Brian Morin: Right. Well, I think that that’s actually a pretty good segue into the very next point in this, Aarde, which is we were calling out three facts or highlighting three areas that we saw big impacts. As you mentioned, one thing that COVID is it put a huge spotlight on your business continuity. Before COVID, you had known what’s really planning for, the what if scenario of all of the stay-at-home orders that might jeopardize this continuity in the contact center? No one was thinking about AI in terms of a critical component to a business continuity plan so they can remain always on you and have the ability to scale up self-service with demand and scale it down as demand decreases.
So now, the spotlight is on business continuity. Before COVID, almost all of our conversations were around cost savings and CX, so cost savings and CX. Those were the business drivers for change. The business driver for change is different now, and that’s been different, I would say, going back to the end of February, where it’s been a business continuity driver, organizations, virtual agents that are perfectly trained and are always on. We did this as an example for your AAA in the area of roadside assistance. We frankly didn’t even consider cost savings as part of the equation. They just needed the business continuity, the ability to scale up and scale down virtual agents to handle emergency roadside assistance.
Now, we had told in the events, “Hey, we guarantee you are going to see at least a two-third reduction in your operating expenses.” They were very candid with us that they didn’t care. That was not their business driver. It wasn’t a huge piece of their strategy. They just needed to ensure continuity of their customers, for sure, enough post-appointment. We did a virtual session with him just recently, where they came to the table and said, “Yes, we got what’s needed.” The CX was on part live agents. But we were pleasantly surprised to find out that we did find those cost savings in our opex while delivering the seaside. So Aarde, in your organization, how much has that spotlight shifted on business continuity and how much of a struggle was this for TechStyle and navigating your way through those rough waters?
Aarde Cosseboom: Yeah. I have a really good example of this, a really good use case of this. So our company is very supportive in the Black Lives Matter Movement, and we actually participated in the Blackout Tuesday that happened I want to say about a month ago, which means that we completely shut down all of our inbound sales and completely shut down our support team. So talk about an event that was… How do we run a business with all of the doors and all of the people not working on a single day?
From a support standpoint, we would not have been able to do it without SmartAction or AI or self-service in place. So while we still shut down, our live agents, transferring through live agents for 24 hours. We still had the AI self-service tools up in place. So members can still reset their password. All of the self-service, transactional things that they need to do with their account cancel their memberships, get the month, anything that’s easy, and they have to do on a transactional basis, they could still do without having to talk to a live agent. So we went live with a total shutdown, with zero people answering calls, chats, and emails for a whole day and just relied heavily on our self-service automation and tools.
If that’s not business continuity, I don’t know what is. I would hate to do that for more than 24 hours or more than even an hour on the day. But the ability to even have that in our tool shed as a tool to lean on is remarkable.
Brian Morin: Yup. Now, the third piece here that we’ll talk about just a little bit, and Aarde, I’ll throw this your way to talk about how can the contact centers think recession proofing their contact center as a strategy. While we have seen a lot of organizations, they put that emphasis on business continuity as the business driver for their AI initiative, we have suddenly seen a new business driver that’s emerged. This has emerged very recently in the last 30 days, where we see this growing driver around recession proofing that we’ve never seen before. There’s an enormous amount of uncertainty in the market right now. How much will the economy suffer until there’s either a vaccine or herd immunity.
For some organizations right now, they’re trying to get out in front of that uncertainty with a recession proofing strategy, not knowing the nuts and bolts maybe of how that needs to come together as much as it’s knowing that it needs to be on their landscape of, is there something they can do, and what can they do. So Aarde sitting in the seat, if someone isn’t an at-risk business, and they’re considering recession proofing as a strategy for the contact center, what are some of the most important steps that you can think of that they should be taking right now?
Aarde Cosseboom: Yeah. So recession proofing is definitely interesting. What we like to say internally at TechStyle is the ability to scale up and down easily and affordably without having to spend a lot of time, energy, and resources and money to do that. So with things like COVID and this pandemic and then eventually coming out of the pandemic into either a recession or an economic recession or the opposite, what COVID has done, it has taken a pretty historical forecasted normal ability for us to say, “Okay. This is the amount of head count we need. This is the goal we need to get to. We’re either going up or going down or staying stagnant.”
This pandemic has thrown a wrench into that, and it has pushed those forecasts either way up or way down. For some people, it’s pushed them up and then down or down and then up within weeks, days, or even months. We’re going to be in this post-COVID environment, where we don’t really have the ability to successfully forecast volume, trends, and the like for customer service or sales teams.
So I think the idea here is since we’re in that environment and we have to plan with our gut and react accordingly without having super accurate forecasting for our staff members and head count and volume, we need to have tools like AI and self-service tools to help hedge the bat.
So the example here is we know that we’re going to go into what’s traditionally our busy season, which is called boot season around CyberWeek and Black Friday. Of course, all retails and e-commerce go through that. But we don’t know to what extreme. Is it going to be 2X, what it was last year within the forecast, or is it going to be half the volume that we need? So with technology, we could leverage self-service tools even more so that we don’t have to hire an overstaff, a call center that has no calls, chats, or emails coming in, or vice versa. We can understaff, and then if the volume comes in, ramp up our self-service tools so that we can contain more contacts via the phone or chat channel.
Brian Morin: Aarde, it reminded us the conversation we were having around this, that when it comes to contact center leaders, one of the biggest risks that they’re having to prepare for right now is if the economy does go through a trough. If there is not a an immediate V-shaped recovery, it puts context center leaders in the position where they have to furlough agents and furloughing them for an uncertain period of time. But when that volume comes back and that volume is going to come back, we all know that volume is going to come back. Now, they’re in a very difficult position. If you have such a heavy reliance on the human skill workforce, you’re needing to scale up, and you’re needing to rush to train a pool of agents, I mean, right now.
Aarde Cosseboom: Yeah. It’s all about the loss of tenure, the loss of people who had the tribal knowledge, who knew how to handle these calls and chats quickly and navigate your systems. So you might be able to retain some of them or bring them back, but you might not. So you might have to have a net new person that you have to train who’s new off the street, who may take a month, two months, three months to ramp back up. So the human capital side of the house, the ramp-up time to train someone is extremely lengthy, costly, and hard. With conversational AI and bots, there is no ramp-up period. It’s a virtual system that is using AI across all of your conversations. It already knows it’s already trained and fine-tuned. All you’re doing is just ramping up the amount of volume that’s going through it.
So the ramp-up time for that is very low to nothing, and it’s a very minimal resource-wise to be able to scale back up when your volume comes back in.
Brian Morin: So Aarde, let’s go from there. We’ll jump into this, which is a really important part of this conversation, particularly when you’re looking at either just over-the-voice channel or you’re looking at omnichannel over voice and then scaling that experience to other channels. When the pandemic first hit, suddenly, the emphasis was on business continuity. We had a rush of organizations realize that, “Hey, our number one problem is in voice right now.” We need you to do something for us in voice. This happened first with the State of Washington. They were the very first to get hit by a stay-at-home order.
When we started hearing from organizations that were headquartered in the State of Washington, we realized, we’re like, “Oh, organizations are not prepared for this.” Because they were needing to do something over voice because they didn’t have agent availability, and they needed to do something, and they needed it stood up within a two-week period of time. It was an education process for us with those organizations and say, “If you’re trying to do something in voice and stand up an application for across different interactions and voice, you do it two weeks, it just is not going to happen.”
Building something where the CX is on par with a live agent over voice is a very complex channel. After you’ve dotted all your I’s and crossed all your T’s, before you even get into the build process, the build process itself is a four-to-six-week process. So if you are looking at voice as part of your recession proofing strategy, you have to understand the lead time that goes into. You’re creating something that’s really going to work for your customers. Aarde, you’re aware we were building something for the national COVID-19 you’ll hotline. They wanted something built in two weeks. We’re like, “Once again, this is a voice channel. That’s impossible.”
But what we were able to do is come up with a channel deflection, where we were able to build something digitally, a chat bot over which rep chat, a very nice experience that we rushed that build, and we’re able to deliver it in two weeks, but it meant deflecting the incoming voice channel over to a digital channel. That’s not how you want to deliver that experience. You want to keep that experience as the channel of their choice. But with the back against the wall, that was something that organizations are doing.
Aarde, I know that you have this conversation. It’s one of the very first things in explaining about conversational AI, coming from the world of an IVR, where should design build in, poof, you’re done. It’s almost more of a product conversation with machines, if this is different, isn’t it, Aarde?
Aarde Cosseboom: Yeah. I love this infographic because to do this right, really, it’s conversations with your customers or members or whoever they are that you have to hit all six of these. So I love the infographic. I usually start at the bottom. So if you’re going to start thinking about a project like this, you start with the project management, then you go to reporting and analysis. How am I going to measure its success? Then training and tuning. How is it going to learn from itself? Is it just going to be stagnant. Is it human learning, AI learning, both, kind of a combination. Then it can’t be stagnant over the years. So you have to understand what the roadmap is to enhance it or upgrade it over time, not only for your business, but also for the natural language processing.
Then the last two are really focused on design. So that’s more strategy and then taking that strategy or that conversational AI solution design and then feeding it back into actual use cases and business outcomes. So it’s kind of like a circular project. What I find happens a lot, and this kind of bleeds into the next slide is the, do it yourself versus go with a partner. The majority of the time, and lots of different factors here, people say, “Well, let’s just build one ourselves.” They usually can’t get past the project management or maybe even just the reporting and analytics side because it just takes too long, and there’s too many resources that you have to hire on board or external resources that you have to pay to help implement that costs tens or hundreds of thousands of dollars.
So I think the solution here, and especially with this COVID environment, because you may find that your business needs this, and it means it stood up before we’re out of the pandemic, and we don’t know when that is. It could be three months from now. It could be a year. It could be next week, who knows? So that sense of urgency and then also the ability to go to market very, very quickly in all six of those different areas is one of the reasons why I would lean heavily on recommending using a partner who’s done it before, not only with other companies, but also has the resource and headcount infrastructure to help support the things that we naturally don’t have in our customer service teams today.
Brian Morin: Now, I know that you’re here at the bottom of the hour. So we’re just going to skip over rebranding slides and move into Q&A here in just a second. But Aarde, when somebody comes from the world of IVR in a matter of a few short days, you design build, and poop, you’re done. You may not even touch it again for months or whatever. Conversationally, AI, it’s very different. A solution requires care and feeding from a team of experts who understand the complete toolset at their disposal from design to build the ongoing operations. Aarde, once you go live, you’ve only started. You’ve only even begun. Correct?
Aarde Cosseboom: Yeah. That’s the tipping point is you have to commit to creating this. Once you do create this experience, you’ve already committed to your customers that you’re going to continue to evolve this and make this a good customer experience moving forward. If you don’t, it creates really interesting negative effects. So it’s almost as if you shouldn’t have even started if you’re going to have to bake it or on only release it and it be stagnant or stale or not the best outcome six months or a year after you roll it out. You’ll get more negative NPS scores and CSAT scores.
If you just release something and let it sit then, if you release something and then continue to evolve it and make it better over time because people will start to expect that self-service tool to do exactly what they want, and if they go to it, and it can’t do what they want six months down the road, that’ll cause frustration and customer friction.
Brian Morin: So if you are listening in, and you are one of those that we want to jump into that docking challenge of handling and automating and voice and then just scaling that experience digitally, you may wonder, well, what kind of CX team do you really need to do everything from the design, to the building ongoing operation? That’s just gives an idea of the type of team that we have curated over the last few years. Each face represents a full team behind them, and it’s what we’ve seen that we’ve had to develop to support the organizations that we manage their AI-powered CX for. We’ve seen similar teams in mirrored who are also doing something themselves, like for instance, like a capital one, who’s chosen to build their entire experience on their own and control that from their end. Very similar.
So we’ll just kind of skip over here, just the very end. We were going to get into some of your results, Aarde. But if somebody wants to bring that up during Q&A, we can certainly get into that. If you are interested in next steps, at least for SmartAction, just see on screen input smartaction.com. You can drop us an email, and we’ll start with you requesting a demo. You’ll call types or see chat interactions in your industry that we’re likely already supporting for somebody adjacent to you and/or an AI readiness assessment, just to understand, what are the thresholds for change where you can expect to get a very strong ROI?
So with that said, what we’ll do is we’ll just come back and see what has come in from a question standpoint. I can’t recall. Is it, Julie. Are we throwing this back to you? Or was there somebody else on your team that we’re throwing this back to?
Julie Lawson: No. No. I have a couple of questions for you from the audience. We have one coming in. Someone’s asking, when it comes to the customer satisfaction of virtual agents versus live agents, what’s the difference to the clients?
Brian Morin: Yeah. How about that you-
Aarde Cosseboom: No, that’s a great question.
Brian Morin: … pass that one, Aarde?
Aarde Cosseboom: Yeah. I get that one a lot. So we do one survey, and we do two kind of different… Sorry, doorbell’s ringing. I mean, we do two kinds of things here. We do a customer satisfaction survey for agents, and then we do a customer satisfaction survey for our self-service tool. What you’re seeing on the screen here on the left click kind of bubbles. I know it’s the animated slide. So we kind of skipped through it. But here’s some feedback that we got from some of our customers. I think my favorite one is… I can’t believe I had the most painless refund process with ShoeDazzle ever, and it was through an automated machine. I didn’t know that was possible. Just something kind of cool to think about. But we measure them against each other, and we have a 92% member satisfaction survey score specifically to our voice bot.
Julie Lawson: All right. Awesome. All right. We have another one coming in. Someone’s asking, how long does a rollout of virtual agents typically take?
Brian Morin: It depends, depending on the complexity of what is being rolled out. It depends on how many interactions you’re trying to automate. Let’s say the most common example would be somebody who’s starting small. They might start with two to three interactions that they want to try automating at first so they could test the experience and go live. For us, we see it’s typically an eight-week process to get all stakeholders involved to agree on what the conversation your flows look like, what the business rules are, your understanding. Because again, things like call recordings. We need to understand exactly how agents are handling calls. There’s quite a bit that goes into the CX design piece from our human experience activist.
Of course, then there’s also training the language models and the attempt. So there’s quite a bit that goes into it. But if all ducks are in a row and nothing delays on the client side, this is typically a eight-week process to build.
Aarde Cosseboom: Yeah. Just to piggyback on that one a little bit, so partnering with a company specifically like SmartAction, they may already have intense or use cases for other companies like billing or collections or inquiry or status on a shipping status, something like that, or no authentication of the member or scheduling of something like a dentist appointment. They may already have that built out. So it’ll speed up the process if you’re partnering with a company like SmartAction versus building it yourself, which can take months upon months.
Julie Lawson: All right. Great answer. We have another one. Someone’s asking, when it comes to implementing conversational AI and virtual agents, would you recommend tackling voice and chat at the same time?
Brian Morin: Yeah, Aarde?
Aarde Cosseboom: I’ll kick this one off, and then we’ll pass it over to you. It all depends. I mean, it depends on the business unit. It depends on the way that your volume sets. Tackling both at the same time might be complex because you might have certain intense that are different based off of channel. People may be more apt to call you for certain things versus chat with you for others. If that’s the case, then you might want to separate the channels and focus on one first, probably your highest contact volume, channel first, which most likely will be found. If the channels are the same intent-wise, so you see kind of the same conversations happening on both, and the conversations are easily deflectable, we call them transactional, then you could probably do both within the same scope. You probably want to launch one ahead of the other just to see performance. But you can design for both at the same time. Yeah, Brian, I don’t know what your thoughts are, if one’s better than –
Brian Morin: Yeah. So yeah. I think that you nailed it, Aarde. There are exceptions here where we may have applications and templates that are essentially already prebuilt. We know that you’re not going to stray far from those. An example might do. Right now, we’ve had a slew of auto dealerships. You’ll come to us to handle scheduling. That’s something for us that we can scale that out over voice and text or chat simultaneously. But for the most part, if you’re doing more kind of like triaged type interactions, which means you’re asking questions, you’re gathering information, and then you’re taking some sort of action, whether that action is you’re putting something into the CRM, or you’re transferring to a live agent, if it’s a little more complex in nature, we’ll build that out, and we’ll shake that out for the first few months.
Then once we feel really solid about the business rules, the guardrails that have been created for the virtual agent, we know we’re delivering a great experience, then we know that’s the time to scale that to a chat because otherwise, you end up having to make all the fine-tuning and changings and all those channels simultaneously at the same time.
Aarde Cosseboom: Yeah. So I forgot to mention one little thing there. It may be daunting to tackle the phone channel first because it may be considered as the harder channel with natural language processing and text to speech and vice versa. Partnering with SmartAction made it a lot easier for us to do that, to go to market with the voice bot. So it leveled the playing field between choosing between voice or chat first because they already have that technology in place. So it was a no-brainer for us, tackle the channel that has the most volume that can be the most successful, which is voice and then do you chat after as kind of an easy deployment out because it’s the same exact flow, same exact intent, same exact… Everything is just-
Brian Morin: Yeah. Same flow. Right.
Aarde Cosseboom: … different channel.
Brian Morin: Yep. Same data, same API.
Julie Lawson: All right. Awesome. We have another one coming in asking, if you’re in an organization that’s never used any type of AI or virtual agent, how do you get started?
Brian Morin: How do you get started? Well, I guess, Aarde, how did you get started?
Aarde Cosseboom: Yeah. Two things. One, do your research. So to check that one off the box, you’re already on this call or listening to the recording. While you’re doing your research, request a demo, listen to some case studies, listen to how the technology would work so that you can formulate it in your own mind. Then the second step, which is a little bit of the harder step is involve your executives, the decision-makers if you’re not the decision-maker. Pull the right people in early so that they know what you’re looking at and that when you go to them with this idea, it doesn’t come from left field. This is what I’ve been researching. This is how I think it can solve a lot of our business problems and unique needs, and this is a demo or an example of how it would work for us.
Get them involved early because executive buy-in can be extremely hard for these types of projects, mainly because most people at the executive level will say, “Well, this might not be important now. Let’s just hold off another year.” You’ll hear that year over year, over year until it never gets approved.
Brian Morin: Yeah. So one thing that we do offer that makes at least a little bit easier for folks to get their heads around if their organization is a good fit for AI automation or not is we do a free consultative call study or chat study if you want to do chat first, where we look at your top drivers into your contact center. We identify how transactional in nature are those interactions that are taking place. Something that can be very complex, and we know we can still handle them others to require critical thinking or human judgment. Then we’re looking at certain volumes. You have to meet the right thresholds for change. Otherwise, the ROI just isn’t there for you to do the design and build across everything.
An example would be Office Depot. When they came to us, they were having a lot of volume coming through on returns. But we didn’t say, “Hey, let’s make sure that we’re containing 100% of our return.” Because in some cases, somebody is wanting to replace the item with a different size. We need to find out, is it in stock? There’s a lot of complicated backend processes that have to be built out to support that, and we looked at, well, if there’s 60% of calls that just want a refund, why don’t we start there?
So somebody’s just looking at the nature of your interactions and identifying how much of that interaction? What is the highest volume path that most people take, what we call that the happy path that’s straightforward and easy to automate. Why don’t we start there? So you do need that level of consultation and then understanding exactly how all of your agents handle every exclusion to understand, what should stay at the live agent, and then what should go with the virtual agent. Then that way, you’re designing and crafting the symbiotic relationship between virtual agents and live agents in your context.
Julie Lawson: Great answer, guys. That’s about all the questions we’re getting from the audience. I also just want to take this time to remind everyone that we’ve been recording this session, so it will be available for you to watch on our site within the next 48 hours. Brian and Aarde, if you have anything else you want to cover for your presentation.
Brian Morin: No, other than to thanks Aarde for… I know he has been in schedule and a busy man, so we just appreciate him joining in and helping sharing his perspective from the front line.
Aarde Cosseboom: Of course. Yeah, and thanks, Brian and the SmartAction team for hosting and putting all the content together. I think, like I said earlier in the presentation, feel free to reach out to SmartAction directly, or if you don’t want to involve them and you want to go directly to me and just ask some questions about AI or how we partner with SmartAction, I would love to connect. You can reach out to me directly on LinkedIn, and the company I work for is TechStyle. So yeah, just happy to be here. I live and breathe AI and conversational AI. Brian, thanks for hosting and bringing me up.
Brian Morin: Great. Thank you so much.
Julie Lawson: Yeah. Thanks, guys. That’s all we got time for today. I hope everyone had half as much fun as I did with this great presentation. Thanks again to our sponsors, SmartAction and again, to our guests Aarde and Brian and for all of you for attending. I’m Julie, and you can find Customer Experience Update on Twitter @CXupdate. Thanks, everyone, and have a great rest of your day.
Brian Morin: Yeah. Thank you.
Aarde Cosseboom: Thanks. Thanks, Julie. See you.