2023 State of Intelligent Authentication and Fraud Prevention

[00:00:00] Matt Smallman: Good afternoon, my name’s Matt Smallman, and I’m the founder of the Modern Security Community. Thank you very much for joining us in this afternoon’s session, where we’re gonna be joined by Dan Miller and Derek Top from Opus Research. Uh, and we’re gonna discuss the latest research findings from their 2023 Intelligent Authentication and Fraud Prevention Survey, which I think has some really, really, insightful thoughts for, for us as an industry, both end users and, uh, implementers also, solution providers of this technology. So, uh, without further ado, we’re gonna jump over and say hello to the team.

[00:00:37] Hey, Dan. Hey, Derek. Thank you very much for, for joining this afternoon.

[00:00:40] Dan Miller: Hello.

[00:00:41] Matt Smallman: Just want to introduce yourselves-

[00:00:42] Derek Top: [inaudible 00:02:17]. Thanks for having us.

[00:00:44] Matt Smallman: … very quickly to everyone who’s on the line. Dan first.

[00:00:48] Dan Miller:

Introduction and background to the survey

[00:00:48] Dan Miller: Sure. I’m Dan Miller. I’m the lead analyst and founder at Opus Research. And over to you, Derek.

[00:00:54] Derek Top: Yeah, I’m, uh, Derek Top. I’m the senior analyst and, and a, a research director with Opus Research. We’re a, a third party analyst firm. We, we look at technologies around, um, security, uh, in addition to, um, self-service and conversational technologies.

[00:01:08] Matt Smallman: Thanks very much. So, so, I, I think the first place to start is just a bit of a background to this survey. So, this is… Is this the second, the third, or the fourth year? I’ve, I’ve lost track, Derek, of-

[00:01:18] Derek Top: Yeah, the, the-

[00:01:19] Matt Smallman: … how many times you’ve run this survey.

[00:01:20] Derek Top: We have done it a few times. I mean, this, the, the, for the purposes of this, uh, for this report, we’re looking at kind of a, year over year. Uh, so the last two years, we put together these, the same kind of questions in the same kind of format to, um, uh, uh, to a group of executive decision makers and implementers, uh, looking at security and authentication. So, um, the survey is, uh, 250 folks. Um, 2, 125 in the UK and Europe, 125 in, in North America, with, with the goal is really to understand how contact centers are looking at authentication, uh, and, and fraud prevention processes. So, we do ask about, um, you know, terms, in terms of, uh, the, the, the digital and voice channels, how, uh, um, the contact centers is looking at, uh, you know, the, the volumes in them. What they’re doing from an authentication perspective. Uh, and then, also from, from a fraud, fraud, uh, prevention to, to thwart some attacks. So, that’s the goal.

[00:02:12] Dan Miller: Yeah. And, and I would only add, and very quickly, that the, this is one year tracking from the year prior regarding sort of a combination around what we call intelligent authentication. We’ve been s- doing surveys of, um, enterprise implementation strategies for voice biometrics for seven or eight years [laughs], as you were saying, uh, Matt. And, and that’s something, uh, to sort of keep in mind as you look at their, the results. We have consciously added, uh, more factors. Um, we regard and, and, and looked at sort of the ancient factors that seem to persist as a, as a bunch of candidates that are more modern are, are presenting themselves.

[00:02:58] Matt Smallman: And, and, and for me, one of the most interesting thing about this year’s survey is that, historically, we’ve done this as a, as a snapshot to take a point in time view as to what people think. Um, but we’ve never really gone back to look at whether that happened or not. And, and this year-

[00:03:13] Dan Miller: Yeah.

[00:03:13] Matt Smallman: … for the first time, we’ve both gone back and said, what did happen to the same, to some of the same people who said what they thought was gonna happen, as well as what, what is gonna happen.

[00:03:22] Dan Miller: Right.

[00:03:23] Matt Smallman: Uh, and we’ve also been able to expand the questions to include some really key questions around how that process is performed today, so we can start to develop some benchmarks for the industry as a whole to see how people are doing. And that’s, again-

[00:03:37] Dan Miller: Mm-hmm.

[00:03:37] Matt Smallman: … blew out some really interesting results, which we will, uh, now get to. So, uh, n-no, no suspense, uh, jumping straight into the, uh, into the presentation.

What’s happening with call volume, complexitity and duration?

[00:03:47] Matt Smallman: I, I think the first thing, and this will be as no surprise to anyone who works in the, in the call center industry, uh, is that, um, compared year over, year over year, in, in most cases, everyone expects call volume, complexity, and duration to increase. So, the age-old death of the call center, death of the agents, uh, is mostly overplayed. Uh, and, but I don’t, I don’t think this will be any surprise to, to anyone. I don’t know. Dan or Derek, did you have any kind of specific observations with, with these sets of results?

[00:04:21] Dan Miller: Well-

[00:04:22] Derek Top: Well-

[00:04:22] Dan Miller: … I think it’s very… Oh, go ahead, Derek. [laughs]

[00:04:25] Derek Top: Yeah, no. I, I’ll just say, like just a, just a note too, like this is one of these things too. Because, we did do the survey in just in the post-pandemic world. And I know we, we’re kinda sick of talking, talking about the pandemic, uh, in some ways. But, but there is something to note there in terms of what we, what we may’ve seen, uh, from, from initial, uh, you know, kind of, d-demands, uh, around, around, the, you know, 2021, uh, and then that first year after. Um, but, but I think it’s just, just something to, to, to note. But, but yeah, no, i-in terms of the, the no surprises, I mean, call volume, um, i-i-s up. And, and I think, that’s, you know, that’s part of, part of what we’re, we’re seeing that the, the contact center is never, is, is always the, the first point of, of return for many customers.

[00:05:03] Dan Miller: And, and we have long-term coverage of what we call conversational technologies. And this is very consistent with, um, with [laughs] what we’ve seen in terms of the staying power of the contact center in the face of growth of digital channels. But, um, and all are growing. That, [laughs] you know, Derek, Derek mentioned the pandemic. The pandemic has lapsed into being the endemic now. And we’re seeing what, you know, what is lasting after there were, you know, ex- uh, extreme growth [laughs] in the use across all digital channels, which exposed sort of issues, um, uh, and weaknesses in the current delivery channels. But then, as volumes has returned to… It didn’t regress to the mean, it regressed to sort of [laughs] the post, to the endemic, um, the endemic state of, of things right now.

[00:05:57] And, um, it’s just pointing out that call volumes, um, for sure went up. And they will stay up to a degree. The complexity of the calls, um, that, that gets really interesting, because there’s, um, there are… You know, as you look at these conversations that companies have with their clients, their customers, um, they are asynchronous in nature, and they cover a multitude of channels. And one of the first order challenges that has emerged is, uh, can you authenticate somebody in a digital channel, and then when they move over to voice, um, do you have strong confidence it’s the same individual using that voice channel? And that’s something that’s going on in the background here as well.

[00:06:42] Matt Smallman: Yeah, I, I completely agree. And, and as, as a, as somebody who’s spent most of their career in customer service, this, this kind of chart is the, is the, is the red alarm chart, yeah. So, volumes are up, duration is up, and complexity is up. And that equals, ultimately, I need more people, and more resources, and better resources-

[00:06:59] Dan Miller: Mm-hmm.

[00:06:59] Matt Smallman: … in order to meet that level of customer demand. So, immediately, I need to go into bat with chief operating officers, with finance teams, to, to secure the resources I need just to continue delivering the current expectations of service, let alone anything above and beyond that. Uh, and I think-

[00:07:16] Dan Miller: Right.

[00:07:16] Matt Smallman: … that’s what leads on really, interestingly, to our, to our, to our next point.

How have priorities shifted?

[00:07:20] Matt Smallman: When we ask people what their priorities for work in the intelligent authentication and fraud prevention space was, in, in 2022 coming off the back of the pandemic, it was really clearly orientated around security.

[00:07:32] Dan Miller: Mm-hmm.

[00:07:32] Matt Smallman: People were worried about the money going out of the door and/or the data going out of the door. And most projects were focused on security. And that has swift, switched wholesale back to efficiency and usability as the priority going forward for this. Uh, and I think that’s, for me that, for me that’s a really, really interesting, um, switch, and, and shows the transition from the set of technologies we’re gonna talk about in a second from being, uh, kind of a security focused thing to something that has real business benefits and opportunities, as, as Dan mentioned. Um, any, any thoughts-

[00:08:07] Dan Miller: Um yeah.

[00:08:07] Matt Smallman: … from you guys on the, on that particular finding?

[00:08:12] Derek Top: Yeah, well, I, I always think the fact that, I think to the point, it a-a-allows some creativity and flexibility in terms of your, uh, current security and authentication processes. W-With, with the idea of preventing fraud or just like shoring up security as the main priority for, for organizations, uh, two, two years ago, um, really does speak to what, what is possible from, from a authentication standpoint. And, and usability, there, there is, this gets back to the whole idea around con-consumer expectations and, and, and what, um, when, when people are calling into a contact center or, or, or, or just going through digital channels, they have a problem to be solved. Their, they don’t really want to go through the hoops of, of security, um, authentication. So, so allowing some flexibility around usability gives it personalization. It gives, you know, it really kind of opens the door for some, some benefits of, of what, um, more streamlined authentication processes can be.

[00:09:01] Dan Miller: Yeah, and when I think about the internal sales job that the, um, contact center folks have to do to bring in new authentication technologies, this is very consistent with what we’re seeing, uh, enterprisewide in terms of investment in technologies that improve the customer experience. There’s a real focus on making things easy for the customer, as well as making the agents more production. And, and those two factors combined, the efficiency which translates to a bottom line sort of savings and the customer experience which maps to things like customer retention, task completion for customers, and that sort of thing, um, we have to be mindful of that. And, and this is very sort of internal politics, uh, correct [laughs], internally politically correct, if you will. There, there should be a focus… I mean, there’s a, there’s a constant balancing act between the two, as, as we’ve mentioned. But, um, the, you know, for, for the coming couple of years, the, the pendulum has swung more towards this, you know, improved customer experience.

[00:10:06] Matt Smallman:

What is the automation opportunity?

[00:10:06] Matt Smallman: And I, and I, and I think when we add these two set, two kinda data points together, plus whatever’s happening in the rest of the economy and in technology in general, I think it’s a really exciting time. When we think about the capabilities of conversational commerce that you’ve been talking about for, for ever since I’ve known, uh, known you Dan, Dan and Derek, about the potential of that technology, it seems that we’re now on the cusp of that potential being realized, the interfaces that those, that those systems provide are now a state of usability and effectiveness that allows people to have the kind of interactions that they might’ve previously had with an agent. Not every interaction, but certainly a fair chunk of them.

[00:10:48] And then, when we look at the advances in the last kind of six weeks, let alone six months, in terms-

[00:10:54] Dan Miller: [laughs]

[00:10:54] Matt Smallman: … um, A, AI, and, um, the ability for that to integrate, and to plug into everything, and connect to the gaps between stuff, that kind of… The second challenge, which we’ve always had in the kind of the long tail of customer services. Yes, we’ve always been able to automate balance inquiries, and position inquiries, and where’s my delivery type inquiries, because they’re low volume, and, uh, the low complexity and high frequency. But that, that enormous long tail of inquiries that’s nearly always needed a, a human to, um, understand, if not necessarily from the customer’s point of view to deliver any particular added value, um, those are now within reach of kind of the conversational, um, solutions that, that we’ve been talking about for, for many years. And I think that’s, what’s really exciting about it is that, that, for the first time, that increased volume, demand, and complexity, we have a real chance of, um, meeting some of that demand with automation, um, that we never previously would’ve had. We’d have to throw people at this kind of problem in the, in the past. I don’t, I don’t know if you agree or, or disagree. It might a, it might be a bit of a extreme position. Dan.

[00:12:01] Dan Miller: Oh, not at all extreme. That, the, um, you know, I was thinking, as you were talking about that, is, is what becomes interesting as these, um, AI, as this AI layer takes over sort of the customer care and the contact center, um, domain, that a bunch of the knowledge that informs the bots that provide better customer service are using personal information, um, to sort of know that person better, and then better match that, um, what they’re doing at the time. ‘Cause you wanna, you know, you, you establish a level of security based on the, the risk associated with the person, the risk associated with what that person is doing.

[00:12:48] And it, and when, you know, this, this conversation culminates in a phone call [laughs] to the, to the contact center, they’re pulling up in, in order to inform the bot to, to provide better service, personal information, and then matching it with stuff from the back office. And it is so important that there be a strong sort of trust fabric beneath it, the thing that’s built around stronger authentication that, you know, even though I’m talking… [laughs] The, the part of providing the convenient, fast, you know, uh, resolution of a problem does indeed [laughs] involve doing strong authentication. So-

[00:13:26] Derek Top: Right.

[00:13:27] Dan Miller: … even though we’re saying, uh, “Hey, convenience is, is prevailing, uh, security,” they’re interdependent. And, and, you know, you have to do both.

[00:13:38] Derek Top: Well, and just, just to, sorry, just to finish that point in terms of our own experience, we just recently did a, a conversational AI awards. And, and a number of the nominees, and in fact some of the award winners, were looking at the ability to kind of have that, that kind of contained, you know, um, you know, the, the deflecting the calls from, from, you know, a contact center based on the fact that they’re authenticated, can, can resolve whatever transaction they need. But also, once they are, and as you were saying, Matt, once they get to a, to an agent, the, the, the, um, the information has been, uh, you know, transferred to the agent that can provide personal, uh, service. And we did see, of course, agent handling time down. But we did really, did also see a number of folks were [inaudible 00:15:53] customer experience does go up based on having the kind of authenticated personalized experience.

[00:14:22] Matt Smallman: Brilliant. And, and just to, just because we have a number of people on the call, if you have any questions you want to ask, uh, Dan, Derek, or I, then please feel free to drop those into the chat or the Q&A feature, and I’ll put them to the, the team at the right point. I think, ha-having, having talked about the opportunity and the promise, I, I think what was-

[00:14:41] Dan Miller: Hm.

How is this constrained by poor identification and authentication performance?

[00:14:41] Matt Smallman: … really interesting for me, for the first time this year, was that performance benchmark, asking organizations, how, whether they, um, ask, whether they tried to identify customers in automation, whether they tried to authenticate customers in automation, and how well they did at that. Now, now for me-

[00:14:58] Dan Miller: Mm-hmm.

[00:14:58] Matt Smallman: … that’s being pivotal to my work over the last kind of almost decade, uh, and what I call the kind of security path of digitalization, which we’ll look at in a minute. Um, but I was, hav-having worked with some very high performing organizations in this space, I, I was completely blown away by how poorly I think we do, as an industry, at achieving these objectives. And therefore, how can constrained advanced AI and conversational solutions might be by the mere fact that we don’t know who the person is or are insufficiently confident that they are that person to actually do the thing they’re asking us for. So, um, just bear with me whilst I find the right slide. So, we-

[00:15:37] Dan Miller: Hm.

[00:15:38] Matt Smallman: … um, we asked cus- We asked people, um, whether they, uh, attempted to identify people with, through automation or with auth-authentication. Uh, and in this chart, you will see 100% of calls kind of come in at the left and the width of the chart. The width of the line is indicative of the volume. And, and a startling 55% of organizations didn’t even attempt to identify the caller, um, before they got to an agent, which I think, for me, in a, in an age where, um, I saw some other research we mentioned in another session, close to 80% of callers to most organizations are calling on their own mobile phone that you probably already have on your records. I, I think that’s just a, a frightening number, just to, just to leave that on the table. You don’t even have to worry about authentication at this point. Just not knowing who they are and being able to make sensible automation and routing decision, uh, was, was quite startling for me.

[00:16:30] But the second kind of effect of those, e-even when we did that, we only managed to achieve it about half the time we asked the question, just under-

[00:16:37] Dan Miller: Mm-hmm.

[00:16:37] Matt Smallman: … we asked the question, which again, um, and we’ll come to this later, I think is testimount to, um, some of the legacy methods that exist, uh, in this. Like, if you are asking people to remember their account number, or the order number, or their customer reference number that you once printed on a letter that you sent them, then their chances of getting it right are, are very low. Uh, and we saw that in the results. So, given that only about a quarter of calls ever get it, get through to an automated identification, automated authentication step, um, it was quite pleasing to see that those organizations that went to that effort did attempt to authenticate most callers. So, only about 10% of what was left ended up with no attempt to authentication.

[00:17:20] But again, when they did attempt, organizations, half of all organizations by, by volume, were unsuccessful, more than half were unsuccessful. And again, I, I think anyone who’s worked in this space for a long time knows that whe-whether or not the user is capable of answering knowledge-based questions in, in an IVR or conversational solution, whether they’re capable of that or not, the user experience is often so poor that they choose not to. And they end up speaking to agents. So, so in, in our model, we saw less than 9% of, um, callers ended up in both an identified and authenticated state before they ended up speaking to, um, an agent. Um, and if we consider that maybe half of those intents generously, generously are available for, um, self-serve, that means that the, the containment opportunity, when we look across the whole of market, was less than 5%. And, and I think that’s just, um, terrible compared to the, the efficacy of these solutions that we see, um, deployed with some of the high higher performing organizations.

[00:18:22] So, for me, this was a real, real eye-opener. And, and I’m really looking forward to tracking this and potentially digging into it in more depth as, as we go through the next few years. Because, the, the opportunity of these technologies that are now affordable, easy to implement, and, and mostly delivery on customers’ requirements, um, is, is enormous. But they’re held back. They’re constrained by the fact that we just can’t get many people to start the journey. They end up in the kind of 9%, uh, bracket at the top. I don’t, Dan, Dan or Derek, I… This is, this is the first time I’ve kind of shown you this chart. Did you have any observations from it for us?

[00:18:59] Dan Miller: Yeah. Well, I would only add, from the, uh, you know, taking the point of view of a, of a customer or, you know, client, or patient of, of, you know, companies use, uh, employing this sort of thing, you know the prevailing experience, you know, puts a [laughs] why to this idea that, that, you know, convenience is now, uh, prevailing. Because, from the customer point of view, uh, without employing the, the automation before you talk to the agent you’re just making the calls longer, and you’re putting the burden on the customer.

[00:19:34] So, if you wanna be consistent [laughs] with your objective of, of placing emphasis on convenience and security, this just shows that you’re doing a, a fairly crummy job right now. And, and, you know, the prevailing model is, you know, if they are using the smartphones, many people just sort of assume that they’re authenticated, ’cause they opened it with, you know, their, their face or, or whatever. [laughs] And, and then, they just hit a, a wall [laughing] when they try to talk to somebody.

[00:20:06] Derek Top: Yeah, and I, I mean, that, that 5% that you talked about, Matt, is just the, you know, that, that just shows the kind of breadth of, of opportunity here around, um, education. You know, I think, they don’t know what they don’t know in terms of the auto-automation opportunities that are out there. Um, and, and you, we brought up a mobile phone a couple times. And, and for the first time, this study and this, in our, in our past couple, um, intelliviews, what we call our kind of larger view of the solution providers in the market, we are looking at net-network authentication. We’re gonna get into this more, but around, you know, the, the, the multiple touchpoints, multiple kind of opportunities to authenticate, identify, customers in that journey, um, is really, I think, the, the, what, what we’re seeing as, um, a tremendous opportunity. And, and that can be, obviously, done, done through automation. So, um, yeah, I think that’s what… You know, that, there, there’s an education that still needs to happen within the market, both, both from a, from a user’s pers- uh, from a customer’s standpoint, and also from a organizational standpoint.

[00:20:56] Matt Smallman: And, and I, I, I’d, I’d agree. I-I-I’ve been somewhat cynical about the effectiveness of some these self-service automation conversational AI technologies for a while. And, and-

[00:21:07] Dan Miller: Hm.

[00:21:07] Matt Smallman: … you guys, you guys all know that. But, but I’ve-

[00:21:09] Dan Miller: Yeah.

[00:21:09] Matt Smallman: … been by what I’ve seen the last 12 months. And I know many organizations are making that investment. And it works for those people that they put through it. But ultimately, they won’t realize the full value of that unless they’re able to put more people through it. And that, that’s what this chart-

[00:21:24] Dan Miller: Exactly.

[00:21:24] Matt Smallman: … really showing us there. We just can’t get people into the start of the funnel.

[00:21:29] Dan Miller: Hm.

[00:21:30] Matt Smallman: So, um, I’ve just had a, a, a comment to say that the chart’s a bit small for people. So, apologies for that. We will-

[00:21:35] Derek Top: It is true.

[00:21:36] Matt Smallman: … send it out afterwards and make sure you can see it. Apologies for that.

[00:21:40] Dan Miller: Oh, you mean there’s letters in there?

[00:21:42] Matt Smallman: Yes. [laughing] There are letters in there. So, uh-

[00:21:44] Dan Miller: Yeah. Oh.

[00:21:44] Matt Smallman: … apologies if it didn’t come out very well on the screen. Um, mo-mo-moving on, I think the challenge underneath this is, um, that, um, most people are still grounded in what we would call traditional identification and authentication methods

What methods do organisations use today to identify and authenticate customers?

[00:22:01] Matt Smallman: . And, and we ask people, what was the primary means of identification/authentication? Now, we actually asking what were their primary, secondary, and tertiary, and what were they investing in, what were they planning to use and deploy. And, y-you more of that in the report. But I think from a summary level, what we see is a vast majo- Well, not the vast majority, but a significant chunk of people are using what we’d call traditional methods for identification. That’s things like social security number, account numbers, stuff that’s easy for you to put into your systems but not necessarily easy for a customer to remember or to, to key back to you.

[00:22:35] A, a fair chunk have moved on to what we call more transitional methods where they are u-using stuff that’s easier for customers to remember. But it still requires some effort on both parts. And then, it was promising to see that actually there are a fair chunk of organizations now, 22% in our survey, um, that are using more modern methods for identification, principally using the ANI or, or call line identifier outside, uh, North America to identify customers. So, so that was really promising, and, and quite a bit of progress since the last time we, we asked that question.

[00:23:06] O-O-On the authentication side, though, the story is, is, is pretty poor. So, the vast majority of people are still using knowledge-based authentication questions, what we call traditional authentication, whether that be transactional, um, or, um, demographic information. Um, we, we even make the distinction that we probably call PINs and passwords transitional, because they at least provide a little bit more security, um, than the kind of information that’s available on Facebook or could be easily researched through, through LinkedIn or, or socially engineered.

[00:23:39] But what was really promising for me, and this the first time this is really even noted as material is that 10, 10% of respondents said that they were now using modern authentication methods, whether that be, um, analysis of network, um, signaling data [inaudible 00:25:31] as calls arrive in the organization, or voice biometrics, or others, as their primary means of authentication. That means the ones in which they are using the most. And, and I’ve been fortunate to work with a number of organizations in this case, uh, who, who are in that situation. And, and it does change the game.

[00:24:11] Now, there’s, there’s a lot more data that sits underneath that. And that’s not to say that organizations aren’t investing in, uh, more modern and, uh, sophisticated methods. It’s just that they haven’t yet reached that level of utilization or adoption that moves it into that primary category. And we do hope that, that will, will increase as, as we go forward. Um, did it, Dan, Derek-

[00:24:32] Dan Miller: Yeah.

[00:24:33] Matt Smallman: … is this, is this, was this surprise to you?

[00:24:36] Dan Miller: Well, I, I think when you start thinking about the numerator and denominator here, um, that e-especially around authentication, even though it’s a, it’s a narrow sliver, it represents a, a fairly, it represents companies with high call volume. So, if you look, if it, if it were, hey, number of calls [laughs] or, uh, dollar value, I, I don’t know exactly what, I, I do think that the move to modern, um, um, as you slice it, and I think it, it is validated in, in our research, is, or in the, um, in the research report, um, that you will find that, that it, there are leading edge companies, probably your clients, Matt, [laughs] that, that are employing these things.

[00:25:20] Matt Smallman: Yes.

[00:25:21] Dan Miller: And, and it is a significant number of calls that are, that are involved.

[00:25:27] Derek Top: Yeah, and I would, um, just also to note o-on the authentication, um, we, we see here, at least in the, in the States, a lot of, um, use of a one-time password, kind of of two, two-factor, uh, a-a-as far as calling in, and then also the repeating of password, you know, that you just received on a, on a separate channel. Um, but that we, that’s within the transitional, we, we think of that. So, that, that’s we think about authentication. That is a common thing here. But it still does have o-o- multiple opportunities for fraud, multiple opportunities for, for poten- for potential man in the middle stuff. Um, and, and it doesn’t provide the true kind of, uh, security that we’ll see around the modern method. So, um, I, yeah, there’s, this is another kind of, you know… And when it comes to authentication specifically, um, the, it, it takes a while for organizations. They’re slow to move sort of, you know, to move from their traditional methods, um, and, and especially if, if they have, um, have, you know, just been using for, for a long time, so.

[00:26:19] But I think we, we, we are seeing it and we will move on to, I think, the, what we see, as the kind of next opportunities around methods. But, um, it is, it’s a slow, slow, slow go here.

[00:26:30] Dan Miller: [laughs] I have, I had an epiphany.

[00:26:31] Matt Smallman: Okay.

[00:26:32] Dan Miller: Um, and, and I call it the airport phenomenon. And, and it has to do with how we are checked in through the security lines, and then what methods the airlines do to [laughs] confirm that somehow you didn’t change identity and you still have your ticket and stuff, when, when you get on the plane. And in the past two years, um, there, there have been a couple instances, one, once in Newark when I was flying overseas where… Or no, when I was flying tra- across the country, where the airline said, “Hey, we’re using a new way to board you on the plane. We’re just gonna… You know, you, you’ll just… We’ll go by using your face.” And then, they aborted midway [laughs] through it. It, it was sort of testimonial that, “Hey, you know, we’re, we’re, we’re a sophisticated company, uh, we think this will be more convenient, and at least we tried.” [laughs] But, but then, then they stopped.

[00:27:26] But, but I, I think more and more frequently people are gonna be exposed, you know, not just in the contact center but, but in the, you know, in, in, in the real world, in real life, um, with, with, you know, better ways to authentication. And, and, um, you know, we’ll see. A-Adoption is slow for a variety of reasons, but the intent is there.

[00:27:48] Matt Smallman: I think, I think that’s, that’s really, Derek, Derek’s point particularly around the, the SMS two-factor, and, and in the full report you can see that broken down into a bit more detail. But yeah, that, that has become the predominant what we’d call transitional method in here, in, in some cases taking over from PINS and passwords. And, and it probably is a, and it probably was an appropriate method for particularly organizations that don’t have very frequent contact with their customers t-to use that, because they have the mobile phone number. Now, now why they didn’t use the mobile phone number to identify them or authentication them when they just call, is, is kind of another question.

[00:28:23] But, but for me, that’s really like, that’s the worst of every situation, yeah? So, you know how painful those processes are, the kind of like, I’m on the phone to you and you just sent me a text message. Now, I have to figure out how to turn the speakerphone thing on-

[00:28:36] Dan Miller: Yeah, yeah.

[00:28:37] Matt Smallman: … and talk to you at the same time. So, have traded off a ton of usability for, in practice, and I think we all know this, very little actual material security, because the phone network was never designed to be used as an authentication mechanism. And therefore, there are, not only are there holes in the phone network and the telco’s management of it, there are also the social en- social engineering aspects, and the fact that we receive so many of these things that we’ll give them out to anyone who asks us, at any opportunity.

[00:29:06] Dan Miller: [laughs]

[00:29:06] Matt Smallman: Therefore, therefore they’re, they’re kind of the, the, the worst of both worlds. But I do see, um, and we’ve talked to a number of companies recently, and I’m sure we’ll be featuring them i-i-in the Modern Security Community, some real potential in, in the, the device as the authenticator. But the device, as a passive, effortless authenticator more than a awkward, cumbersome, kind of workaround authenticator that we see the SMS OTP.

What is the potential for modern methods such as Voice Biometrics, Network Authentication and device based authentication?

[00:29:32] Matt Smallman: A-and, and that leads nicely on to the discussion about potential. Now, now we don’t have a slide for this today. Unfortunately, you’ll have to, uh, you’ll have to reach out to Opus to get a full copy of the, the report to see the breakdown of this. But, for me, this was really interesting for me as well. Like we, we look at three or four key technologies when we think about modern identification authentication.

[00:29:54] We look at network authentication for fraud prevention, which the use of the signaling data associated with the phone call in order to search somebody’s identity or identify anonymous behavior. We look at voice biometrics, which many of you familiar with, the use of, uh, an enrolled voice print in order assert that somebody is who they claim to be or to detect somebody who is known to be fraudulent. And then, the use of a mobile device, either, um, as part of network authentication or to actually initiate a secure call to, from a, from a encrypted devi- from a, from a secured app, to make a call from that through to the enterprise and take all of the security of the device with it in that interaction. And, and those are very different technologies. And which one’s applicable for your organization will vary quite significantly.

[00:30:40] We see network authentication being really, really applicable for organizations that have reasonably low frequency of contact with their customers, and certainly lower risk transactions. But that is an enormous amount of companies. We’re talking about telcos.

[00:30:54] Dan Miller: Yeah.

[00:30:54] Matt Smallman: We’re talking about insurers. We’re talking about healthcare. We’re talking about, uh, utilities. I may, I may’ve just said that, sorry.

[00:31:00] Dan Miller: [laughs]

[00:31:00] Matt Smallman: And, uh, when ask people like, “Do you have the data about your customer’s mobile phone numbers,” um, a huge proportion, the vast majority said, “Yes, we have the data, um, and therefore we could both identify and authenticate them if we had t- uh, access to this technology.” So, I think that’s really promising. Uh, the fact that it’s like the primary means of authentication for less than 1% of our, our vast sample is, um, somewhat, somewhat distressing. But we’ll, we’ll see that evolve over the, over the next period.

[00:31:27] The, the second category was voi-voice biometrics. And really, with voice biometrics is about frequency of call, not necessarily-

[00:31:33] Dan Miller: Mm-hmm.

[00:31:33] Matt Smallman: … just in, uh, in year, but in relationship. How many organizations have a relationship with their customer that goes beyond two or three interactions? A-And, and again, well of the majority of respondents have interactions that makes it worth having the consent conversation, which we’ve talked about in, in previous sessions.

[00:31:52] Dan Miller: Mm-hmm.

[00:31:54] Matt Smallman: And finally, when we look at the kind of secure call, mobile app features, again, uh, a huge chunk of organizations, um, believe that they have their, um, customer accessing their mobile apps, and therefore they can enable this kind of secure calling. So, the potential opportunity for these, this combination of technologies, what we call modern, modern security, um, and intelligent authentication, is, is pretty significant. And I’m, I’m really pleased by the results of our survey and that we’re finally able to start digging into kind of the potential of some this as opposed, as opposed to the snapshot that we may have had-

[00:32:26] Dan Miller: Mm-hmm.

[00:32:27] Matt Smallman: … um, in the past. I don’t know, Dan, you, you’ve been following voice biometrics for like, uh-

[00:32:33] Dan Miller: Ever.

[00:32:33] Matt Smallman: … probably a decade longer than me, at least. So, uh, [laughs] a-any thoughts about how that, how that’s evolving?

[00:32:41] Dan Miller: Um, well, it’s, it’s, it’s been pleasant to see that, you know, the, the real toggle has been around passive authe- or passive, uh, enrollment, and that there has been, you know, greater success in recent years, you know, especially in the financial, among financial services companies in getting that, um, voice print, um, uh, captured. And sort of a nasty term there. Um, I, there’s, there’s still speed bumps, uh, around regulation of biometric information. Uh, but I think some, uh, some case law is being established where, uh, you know, the, the companies that are conscious from the beginning about, uh, being compliant, uh, and being respectful [laughs] of privacy laws, uh, are, are having some success in, in, in getting the voice prints enrolled, and then having seamless implementation take place so that, in many cases, y-you know, somebody at that point where they call the contact center are told, “We just authenticated you with your voice,” and then [laughs] and then it moves on.

[00:33:52] So, so some of the protocol, some of the workflows are, are getting mature. Yes, it’s, it’s considered modern [laughs] compared to the knowledge-based stuff. But more and more, it, it, you know, it is, it is, we’re seeing significant implementations that show it can be done, and it is being done. So, that, that’s cool to see.

[00:34:12] Derek Top: Yeah. I, I would just add in terms of the, you know, we, we always talk about kind of a, a combination of factors, and, and, and kind of a segmentation of your, your customers in terms of what level of risk and security you’re, you’re providing. And each of these technologies provide, I mean, around the network, or voice biometrics, or, or some of the secure call stuff, um, some, some, some priority level o-of, of, uh, security. And so, so I, I think, you know, the, i-if you combine these in, in with your multiple factors with your current authentication processes or security methods, there, there is an opportunity to really kind of, uh, I think take advantage of some of the opp- some of the technologies that are there now. Um, a-and that’s, you know, I think going forward, it’s, it’s not gonna be a, you know, there’s not a single solution for every customer coming into the organization, so I think, that’s, that’s something to look for i-is a, i-is a, is a combination of these factors.

[00:34:59] Matt Smallman: That’s, that’s a really good, really good point as well, actually, the, the comb- the combination of them, using them, using them together, um, and, and separately depending on the nature of the, the customer. And again, we, we have some, some case studies. I think we saw Andrea Ayres from Lloyds Banking Group talking about exactly one of those, uh, earlier in, in our season. So, yeah, look back on that one if you haven’t seen that.

[00:35:21] Just trying to bring this to a, to, to a, to a wrap. I, you know, ask if there are any questions on, on the call before we try and summarize. I think we’re, uh, I think we’re pretty much, uh, out of questions on our, on our call today. So, I’ll, I’ll ask probably Dan and Derek if you’ve got any final thoughts or points before we wrap up.

What is the state of cloud adoption and does it help realise these opportunities?

[00:35:45] Dan Miller: Well, I, I would say, um, that s- uh, we expect in the, in the next survey that we do, uh, to see that there’s been an acceleration in, in deployment of the modern technologies in combination that might largely be driven… And, and we didn’t talk about this, um, some of our early questions were around cloud migration. And I think what we’re finding is that, you know, there’s been a significant investment by firms that always said they won’t, for privacy reasons, for whatever, uh, they weren’t gonna move infrastructure to the cloud, their, you know, for telecom, for customer data, all that sort of thing. And they didn’t until they did.

[00:36:29] And we’re at that point where, um, stuff’s kinda stabilizing in the cloud. Th-there used to be like a, a real extreme up and to the right pattern for investment there. It’s kinda leveled off, reflecting the fact that many of companies are there already, and they’re doing more stuff. Um, and this preoccupation with bringing this layer of in- uh, artificial intelligence into there, um, has, has kinda raised enterprisewide awareness of the value of, of putting money into cloud-based resources. And when you’re doing that, um, you’re, you’re gonna find that it’s easier to sort of cobble together solutions that include the, the modern authentication technologies that, that we’re talking about here.

[00:37:15] Derek Top: Yeah. And I, I’m, I’m curious to see going forward, is, is that kind of that, the, the usability versus security, you know, usability and efficiency versus security. We, we see this balance. And this has been an ongoing, you know, con-conundrum for… Not a conundrum, it’s just a, it’s part of this challenge is understanding when to provide pure security, when to provide more, more, um, allow cust- better customer experience or usability. With conversational AI technologies and more self-service coming in the world, there’s expect-expectations for understanding I, I am who I am. And then, once I get through an automated system, you should know exactly what my, my question or problem is, um, and resolve it quickly. That, that l- that type of expectation is gonna be there cu-customers. At the same time, let’s be honest, fraud is not going away. There’s multiple opportunities for fraud more and more. Um, this, this is, this, this, this balance between security, uh, and usability and efficiency is, is an ongoing one. And I’m, I’m looking forward to seeing what, what next year will bring.

[00:38:08] Matt Smallman: Yeah, some, some really good points. I, I think, Dan, on, on the cloud point particularly, that, that is probably the pivotal moment that we, we forgot to mention. And again, you can see more in the report. That’s, it’s now easier and cheaper. Like so-

[00:38:23] Dan Miller: Yeah.

[00:38:23] Matt Smallman: … so instead of a, a million dollars and 12 months to implement this technology, we can throw it to a developer and for a couple of hundred dollars of transaction costs we can figure out whether it’s gonna work for us or not. Uh, and the same goes for those AI technologies and the, the right combination of those. Because, when we asked people in the survey, the vast majority are now on these cloud platforms. And their focus is not moving to the next new, shiny thing. It’s figuring out how to make the most of it. Many of the, much of the effort associated with cloud migration has inevitably been on like for like. Like just make sure-

[00:38:58] Dan Miller: Mm-hmm.

[00:38:58] Matt Smallman: … my agents can work from home and that the service is not degraded. But inevitably, the pitch decks that got the investment internally included lots of stuff around how it’s gonna make use more efficient and create better value and better customer experience. And now, organizations seem to be turning their, their hand to actually look at that. So again, some really interesting insight in, in the report, which, which leads on really nicely to, if I just, uh, jump, jump to the next slide. So, the report itself will be available next week. You can find it at, uh, opusresearch.net. Or if you want, uh, to talk about a copy or licensing a copy, then you should contact pheadrick@, uh, opusresearch.net. And the address is, uh, all on the screen, and they’ll be on the, on the replay section after this call has finished.

Summary and final thoughts

[00:39:45] Matt Smallman: So, just recap from my perspective, um, this was a, a fascinating survey. And for me, really, is that point where the penny seems to be dropping. Like, all, all things are coming together. First off, we have, uh, a drive for, well, we have the reality of increasing demand, increasing complexity, increasing duration, in an economy where there will be increasing cost pressure. But with a set of technologies in conversational AI and the rest of it, that can actually, actually realize a lot of value and help with that efficiency challenge.

[00:40:19] At the same time, we have many organizations, uh, having completed their migration to cloud contact centers and are now focusing on mi- focusing on optimizing those and realizing the value. But we still haven’t necessarily solved the identification/authentication challenge. But the incentive to do so is definitely there now. Because, if we don’t solve that, then we can’t realize the full value of those kind of investments. And we can’t meet the challenge of increased demand, increased complexity, without adding significantly more people to our organizations.

[00:40:51] So, this was a really interesting, uh, year for the survey. Uh, and as I said, next year we will track whether what people said would happen did happen, and that’s what they did.

[00:41:01] Dan Miller: [laughs]

[00:41:01] Matt Smallman: And what they think’s gonna happen next time. And, and hopefully, you, you guys will join us in a similar sort of timeframe next year, and we can see whether our, our predictions and procosta- procrastinations, pro-prognostications [laughing], prognostications I think is the correct word, were, were actually correct. So, uh, thank you very much for joining us, everyone who joined the call today.

[00:41:22] Dan Miller: You bet.

[00:41:22] Derek Top: Yeah, thank you.

[00:41:23] Matt Smallman: And, uh, we’ll see you again soon. Thanks.

[00:41:26] Derek Top: Thank you.