AI's Impact on A2P SMS Industry

Hello everybody and thank you for joining us today. My name is Nick Millward. I'm an advisor here at the Mobile Ecosystem Forum or MEF. So despite recent challenges, the A2P SMS industry is poised for growth with 2.2 trillion messages forecasted for 2024. Today, Myself and two industry experts in Simeon and James will explore how AI-driven solutions can revolutionize pricing, enhance services and boost margins. So let's unlock the hidden potential of A2P SMS together. So I'll come to Simeon first. Hi, Simeon. Can you please say hello and introduce yourself? Absolutely. So yeah, I'm Simeon Coney. I am responsible for business development at Aenea. So I work a lot with our customers right across the globe, regulators and industry analysts, really looking at trends in the marketplace and how we can help with our solutions. Excellent. Thank you very much, Simeon. And James, it's great to have you with us as well. Could you please introduce yourself? Yeah, thanks, Nick. So my name is James Lasbury and I work for Global Telco Consult, GTC. We are experts in messaging and we help our customers do messaging better. But we also now work in mobile identity and other services. So effectively, we work predominantly with MNOs, so the operators globally, but also the biggest platform providers and enterprises across the markets. Prior to that, I worked for Telefonica and also Vodafone. and Vonage just after Telefonica, probably worth mentioning. Excellent. So thank you, James. We've got a very similar background, except for the Vonage bit. So yeah, I'm sure we'll have a great discussion today. So let's start, Simeon, if that's okay with yourself. So just, if you can just give us your general view on the state of the market for A2P SMS. Yeah, I think we're in interesting times, Nick, I really do. I think we've got a lot of external forces happening on the market as well as internal generated ones that are meaning we're probably heading for a very interesting time over the next 12, 18 months. I mean, certainly we've seen the pricing changes that we commonly talk about having a very significant impact to the market. We've seen brands becoming more aware of not only the price points, but the value that they're receiving from messages in various markets and then potentially tailoring what type of communication they're doing based upon that price point. But also compounded with that, we've seen like a wave of, for example, behaviors like fraud through AIT occurring in the market as well. But yeah, not only that, the growth in rich media as well on some of those other bearers, I think it really means that we're going through quite a transitionary period right now. And James, yourself? Yeah, I'd look at it slightly differently, as in being in the market for some time. I agree with you, Nick. The market is in growth. That's a fact. However, if you break it down into the component parts, if we look at the biggest part of messaging today, which is SMS, that value chain is broken. at the moment and trust is not there anymore. And what I mean by that is where it starts messaging from the consumer, from the enterprise to the MNO who delivers effectively that message into the consumer, that part in the middle where the aggregators and the platform providers sit, the trust across that ecosystem is really being stretched now. The impact of that is that some of the biggest brands in the market are choosing now not to send messages in specific markets due to the cost or even the quality of service. And this is having a direct impact in terms of other channels. So the reason the market is growing, because if it was if it was just SMS, that is now plateauing and predicted to go down. WhatsApp. And RBM, obviously RCS, is coming through now. And that is driving the growth. And the topic that we're covering today couldn't be more pertinent because a big part of the growth there is driven by bots and chatbots. So I think this is a great conversation point. I look forward to discussing it. Right. Yeah. So I think from what I'm hearing, the general trend is that the SMS is kind of peaked very much now. And in certain certain markets, it's declining. And I think, as you say, James, through the through the trust element is where we see most of the challenges right now. But if I can stay with you for now, James, I mean, just, you know, in terms of A2P SMS, there's a lot of. challenges that we are navigating around enhanced traffic, abuse, increase in regulatory requirements. So, I mean, how are the providers tackling this or are they not? Could you just give us your views on that? Yeah, it's an interesting topic, this, because every single provider in the market is very specific on how the brand should protect themselves. So that is stated within... every service provider's website. And the reason that they state that and now provide tools for protecting their customers is because the regulator can implement, basically, they can enforce stops of service and huge fines within the M&O community, which are automatically passed on to the platform providers. So that's written within the contract. And that now, over the last... probably two to three years, we're now seeing platform providers, because of smishing and some of the fraud issues, actually being facing high fines, but also being barred in terms of the services that they can provide into certain markets. So if we talk about kind of, is AIT new? How are the consumers being protected? That's always been there. Are we facing a change in that? The answer to that is no, Nick. What we're facing a change to is how we start to deal with these issues, AIT and the high prices, because effectively brands are moving away. And it's taken that for action to really take place in this market. And the way that we solve some of these problems are through kind of AI and machine learning, which is brilliant because platforms like... Simeon's that we obviously have at our disposal today are some of the ways in which we can block and deal with these issues. For example, I've worked with Simeon both within my past companies, Telefonica and Vonage, and had good partnerships in place to help kind of stop some of that abuse. And that's the key here, Nick. If I can come back to some of the solutions in a second and the products. But Simeon, if I can... Just come to you. I mean, in my position at METH, what I see a lot of is everything right now in the mobile world. And if there's two things that are, if you like, trending at the moment, one is RCS, which we're not here to talk about in great depth today. But the other area is definitely around fraud, anti-fraud, inflated traffic. But from what I'm hearing from all quarters of the industry or all... divisions of the industries, that's rising and it's continuing to rise. So could you just give me your view on that? We'll talk about the products and the technologies in a second, but just give me your view on what's happening in the market, please. Yeah, certainly. Well, I think the first thing is there are more points in that value chain who are getting better insights and understanding on what is actually being carried. I think absolutely, James, I agree with you that sort of... the what is being done hasn't significantly changed, but I think maybe we're now getting a better insight on the scale and the speed of change of various things within that, because we have better ways of sampling or inspecting or gaining control and understanding of what's happening in there. So yes, I think what we're seeing now is that people are able to better quantify things like the different types of fraud. Things like the growth in AIT, I mean, certainly we are seeing not only changes in volumes, but very quick changes in techniques and behaviors. Again, you know, to the earlier point that there's a lot of external factors, you know, price points are changing so quickly in so many different markets, and the brands themselves are responding quickly to that, that we're seeing the frauds also being as dynamic as those two external forces. So- You know, what was true three months ago is not true today. It will not be true of what's happening three months down the line there. People are capitalizing upon those dramatic changes to be able to either exploit for commercial value in terms of a competitive advantage or exploit from a fraud perspective to gain an advantage over maybe their other businesses. It's both positives and negatives, but you'll say that there's a great deal of change and the insights that people are gaining from those are helping them better understand what's actually happening there. So if I understand correctly, there's the artificially inflated traffic that's currently happening, which is being driven by price rising. And again, please correct me if I'm wrong. I'm playing a bit devil's advocate here as well. But so my understanding is that the traffic's being inflated because the price is rising, because it can basically enable the fraudsters to get rich a bit quicker. Is that is that completely true or is there a bit more to it than that? I think that's certainly one of the major factors. I would say another is the ability to programmatically generate large volumes of traffic, often through unsecured interfaces. What we've done in the ecosystem is enable brands and services to generate automated messages, but those services themselves may not be fully protected in the way that people can access and invoke them. So we're seeing a whole range of different mechanisms and techniques being used to generate AIT. And often that does align to markets with high price points, but not exclusively. I mean, we are seeing it sweep across other territories where the price point is lower, in part just because there are brands operating in those markets who can be exploited and utilised to generate that AIT. Or in fact, people can use AIT not just for commercial financial gain of generating a message to get paid for, they can use it to generate an impact to their competitors'business. So things, for example, like conversion rate poisoning that we've spoken about within the meth community also has a consequential impact, even if it doesn't have a direct financial outpayment. It does surprise me ever so slightly. I just, I... I was there at Orange when it launched. That was back in 94. And so it's probably giving you an indication of my age, which is a little bit scary. I don't know how it quite got to this point. But it does surprise me over 30 years, we've got this, if you like, fairly old now in technology terms, technology that's being exploited even more. And it's becoming such a big thing in the industry, you'd have almost thought it would have gone the other way. So yeah. Again, I just find that quite an interesting scenario, really. But let's talk about kind of looking into the future a little bit and maybe talk about some of Aeneas'technologies and products that you have. And the buzzword obviously out there at the moment, if you look at any startups or any, I guess, subdivision of every business now, we're talking about AI and machine learning. So could you just tell me a bit about... an ear's kind of role in that? You know, specifically, what products do you look at that tackles these types of things like fraud? Yeah, absolutely. So, I mean, we've been using machine learning for over a decade now. And in fact, when we first started using it, we were very much using techniques that were looking for what we call lexical similarity. So if we think about, let's say, for example, a spam message, what was happening is spammers were saying, okay, well, This one campaign is blocked, so let's change word. You know, rather than you've won a prize, it's like you've been selected for a prize. So we were using machine learning techniques to catch those variations very, very quickly and also look at the mutation that happened within those campaigns. So that was a system that worked really effectively for a long period of time to catch that type of investment that spammers were making to try and stay ahead of. sort of simple keyword-based and regex-based firewalls. What we're now doing is actually using the new wave of large language modeling to do not just lexical similarity but actually get to the meaning of the message with semantic similarity. And this is where it gets really exciting because lexical similarity, I say, it's about looking at things like word sequence and word replacement, whereas semantic is very much about the meaning of the message. And with the large language models that we now have access to, it's given us the ability to abstract from the actual language being used. And again, we're talking here about the SMS marketplace. It's a global market. And what's fascinating is the array of languages that are in active use by so many major brands. You know, people want to communicate in their language of choice. So as a fraud protection security firewall, it's our job to be able to deliver that protection no matter what the language is. I mean, we today cover a quarter of the world, so we're encountering many languages. And when machine learning really comes to play, it enables us to have that abstraction of language and really get that understanding. So from not only a fraud perspective, we can handle now these campaigns that might jump from language to language where the fraudsters are using translation software. But also we can use these same capabilities to gain an insight to the nature of communication to actually help in other not only value add areas, but also things like regulatory compliancy. where it's again about the principle of the message rather than the exact content that the regulators are getting concerned about. And I guess beyond the whole, I guess, area of fraud or anti-fraud, the applications for that type of, you know, view in the message language, the semantics, you know, has a whole raft of positive connotations where Customer care can manage their customers better potentially, or it could be a sales team could know if somebody is actually interested in buying the product that's being sold over a text message or over a mobile message. Is that what you've seen as well, James, I guess, beyond the whole fraud piece around, if you like, AI and machine learning? Yeah, so for people that don't know our space really well, and I know we all do, it's probably important. to state that messaging, the reason that we have fraud and smishing is because messaging is so successful, right? It's been used in every market globally, and we'll see that evolve. And if we look at, say, for example, the strategy implemented at Telefonica Group across all of the operating businesses, so from Latam to Europe, it was very simple. It started with securing the market, then monetizing, and then... evolving the service. And effectively, what we're talking about here is more effective security of the service with the products that Simeon has brought in. And if we look at it, kind of stretch that out to the MNOs and some of the other services, I know Aenea have been extremely successful in terms of obviously the number of global operators that it works with today. But we are still in a position whereby not every mobile operator has effective security tools. And if we take this back, Nick, to my first point around trust, unless we get the full visibility across the services that are being provided so that the MNO can make the right decisions, the enterprise doesn't have to ask for the tools. It has the visibility around what's going on with its customers and can act in real time. That really then, that changes things because this isn't a question on whether messaging is a successful channel. It is. It's scaling. The issues that we face today are around bad actors and those generating fraudulent activities can't be tolerated in any way possible. Artificially inflated traffic, that's just to inflate revenues for a number of bad actors. So in summary, Nick, I think you could almost flip this on his head. We face a really healthy future if we can solve some of these things. And it's brilliant hearing the way Simeon and his business is adapting. But this isn't new. So I think Gartner predicted that 85% of enterprises would use AI in their customer communications in 2016. And that was their forecast actually by 2025. So that's a year away. It'd be quite nice study looking at whether that, how real that was separately, but effectively AI and machine learning is being embedded in all kinds of communications. And we can even look at some of those players in the value chain, for example, Twilio, Flex, a core part of the service is AI and machine learning. If we look at GMS, one of the biggest providers, in messaging today globally for international messaging, they're effectively pivoting their business now towards AI to manage the customer journeys more effectively and really produce a brilliant customer experience. So I can see your dog just coming around your right shoulder, which is nice, Nick. Hopefully he or she can contribute a bit. I'll need to reposition my camera so you're not distracted. But at least she's not snoring anyway. So that's a good thing on a live podcast. So would it be fair to say that some of this, you know, we're talking about AI and the products in there. But I'm. A large part of this is about visibility as well and opening up that visibility because you've got a number of players in the value chain, as it were, in the ecosystem, I suppose, that are potentially forwarding a single message. It hops through various different companies and through different routes. Is visibility key to this or is it AI that's going to fix it? Do we need to show where things are going wrong? to everybody or should it be should it just be kind of nipped in the bud through this through the technology yeah it's it's an interesting point um simeon they if i give a context in terms of mno then you can give the context the security provider and and nick it's it you you know this as well messaging is incredibly successful but isn't invested in hugely by the mno it's seen as a channel that was growing 15 to 20 percent year on year largely by itself and therefore it didn't get the capex investment into the automated tools to manage the channel effectively so what you had is limited resource with largely manual tools which effectively we're talking about the wholesale shift in that which simeon's capabilities can deliver simeon i'll let you um yeah thanks i think you know what what ai enables us to do is on the visibility point, break apart the fact that it's just a message and go, okay, what is that message trying to do? Who's it coming from truly? What are they trying to do? And with that, you can then start to align it to the value. I mean, I think if we look at other channels, things like WhatsApp, to my mind, the great thing they've been able to do is do use case. based pricing. They've actually said, what is the nature of the engagement that a brand wants to have with their customer? And let's give the value of that conversation to the nature of that discussion. So it's not priced per message. What we want to do is get an outcome. We want to get somebody who's activating a service. We want to have someone who's getting the appropriate level of support. We want to have somebody notified about a service interaction. And I think this is where we see, again, machine learning. helping and gain an understanding of exactly what the transaction is and maybe moving the market away from this sort of commodity unit based price, which only varies destination operator, destination country by country into something that is far more aligned to the value that a brand and a customer is actually trying to achieve. So on that note, in terms, because I would expect the majority of people listening to this webinar are going to try to understand one or two things to understand this industry a bit better and understand our views on this. But the second thing is, and I hear this all the time is, what opportunities are there to monetize these types of technologies? So we talked about what it kind of almost does to kind of try and stop the bad actors and the tools that you have in place to be able to do that. But on the flip of that, how could we use this type of tech to to help enhance the monetization. If you've got an answer on that, Samir? Yeah, absolutely. So that's the key thing. A firewall isn't an inspection and a control point. The fact is that it's been used for fraud and security to start with. But to find new fraud, you have to gain an understanding. To work out what something bad is, you have to understand the context of that and you have to know what good looks like. So yeah, a firewall is an ideal control point. to be able to gain these understandings to say, okay, what is the nature of the transaction? I mean, James, you were talking about regulation as well earlier. This is where firewalls are no longer security points, but they are also regulatory and compliancy points. But off that same back, if you're doing those two functions, why not now start to leverage it for these other value-add services? I mean, I mentioned WhatsApp. I mean, a great example is now starting to profile traffic in our customers'networks and then doing a comparative analysis to WhatsApp. It's really interesting looking at the breakdown of, you know, what's notification, what's promotional traffic, what's one-time passwords. and then applying a WhatsApp pricing model to do the what-if question of, is WhatsApp cheaper or more expensive than SMS? You know, when you're just doing a unit-based price, it's really hard to try and work out that apples-to-apples comparison. But with AI, you get that ability to profile the traffic. And I'll answer the question now. In a number of markets, SMS is actually priced below the price point that WhatsApp is achieving. So this is where we see there's a real future in using AI to gain this profile, to maybe evolve pricing models and again, start to align traffic to the value. And to me, it seems like, I mean, ultimately it's the brands or the enterprises or the businesses out there that are communicating with their end consumers that are paying for this. And my experiences have been they don't, especially if you're talking to, say, a procurement team, they don't get that. they just see the unit price and is that your experience as well because certainly it's been it's been mine that you you're almost positioned a price per message sms versus price per message on some other some other channel like rcs or whatsapp and it's really only until you can demonstrate maybe the roi on that so if you're selling i don't know if you're a hardware store and you're selling a chainsaw or something can you you you're increasing the conversion rate you two times through one channel versus the other. It's only at that point then that you can sort of demonstrate the cost comparison. I mean, have you experienced that as well? And does AI help with that? James, I think I'd look probably more to you as a change support expert. Cool. So here's the thing, Nick. You'll know on the operator side, the cost base is fixed. So the cost is known. And I would challenge your view to it or your point to a certain perspective from is it the enterprise that doesn't understand? Or is it the effectively the partner that's been conditioned into understanding the cost base, not actually communicating with the customer in terms of the value that can be delivered? and really understanding and helping the customer understand the difference between an authentication to a marketing message to something that could be a kind of conversation in the future. From my perspective, managing kind of the carrier side of a platform provider, where the job was to effectively manage the costs throughout the business on an ongoing basis, it was from a routing perspective, Nick. That was such a huge part. If you look at it from a platform provider, it's almost 80% of their cost when it comes to ATP, right? So when flipping that into a customer communication, when the cost is such a high proportion of the actual kind of, of the service that's been provided, the way that the services have been provided today is almost on a trading basis. So they'll deal with a customer looking across a number of markets. and they'll aggregate the price point to give them something more effective. What we're now seeing and what the beauty of AI is actually communication solutions coming out. We touched upon GMS. Their strapline in terms of creating a brighter future for their customers is all about using AI. Twilio has always been at the forefront of this area. They've always had kind of developer tools, but very much. in terms of solutions to the customer. It's really easy because America was their biggest market and then they've expanded outside of that. The other aggregators that have served the other markets have been predominantly focused on how do they control the costs of their service across multiple markets. And I think that's a really important point to understand because sometimes on the M&O side, you don't see that. Sometimes on the enterprise side, you don't get that. It's almost all of that information. is with the kind of aggregator or the platform provider. And this is where another area, Simeon's tools can really help in terms of the effective routing of the messages globally and the transparency of that. Imagine Nika in a world whereby we wouldn't have to talk about hops. You just knew who was delivering the message and then you could see the conversion and then you could decide based upon that who you used. um thereafter i mean we're we're talking about um some pretty you know, groundbreaking stuff, I think here. And wouldn't it be great if, as a brand, you could almost, you know, in real time, see that the cost of a message or something that's being delivered securely and the benefit that that's actually giving you. And just maybe a silly example of that is, I mean, majority of SMS today is one-time passwords from what I see, certainly A to P. But that has a value associated with it. So if... you know, the mechanism through a text message isn't slick enough so that that person is logging into the website. And then let's say you deliver that same service OTP through another channel or some other mechanism that just makes it easier, makes that person maybe log on more, which then has a knock-on effect to drive in customer engagement, customer loyalty, or even some sale of some description through a retail website, then wouldn't it be great if you could demonstrate that through the power of AI. Personally, that's the sort of thing if I was heading up a customer care or an IT team or I was the CMO of a company, that's the sort of thing I'd probably be looking for. So again, I don't think we're too far away from that, from what we're discussing here. Nick, I can just jump in on that point. So to give you a real data point in terms of how AI is helping our insights in the industry. What we're now seeing is one-time password is no longer the dominant traffic. I mean, we know that it was going to be affected by the increase in price points and the rise of things like authenticator alternatives. But using that traffic profiling techniques, we're now seeing that actually notification is now the dominant traffic type in A2P flows. One-time password still sitting at number two and then promotional traffic actually showing the greatest level of growth. in the segment there. So it sounds like brands are starting to think about how can I use promotional messages to sort of re-engage my customers. And as we all expected, the one-time password traffic starting to drop, but actually notification, not only by traffic volume, but also by the number of campaigns, the number of distinct campaigns also is the largest segment now with the highest growth. So again, I think that points to a very, very positive future and the outlook of... how brands are using messaging. Yeah, that's really interesting. Again, that's probably my, you know, naivety at the moment around SMS that, you know, most of the conversation seems to be around the one-time password. But if indeed it is notifications, it is promotional messages, then certainly with the power of AI and other maybe richer channels that are coming to market, you know, we should see a lot more benefits of using those. And I think Stats that I've seen from a couple of different research companies, and I'm sure you know the names of them, which I won't mention here, but they're all saying that the number of enterprises, the number of brands is very low compared to the number of brands that could be using these types of technologies. So, you know, if the messaging providers can really expand the portfolio out to more businesses and we get more usage of these types of technologies, then, you know, I think. again, bringing AI into the conversation and maybe even rich messaging, you know, you become even more successful with this, with this area. So let me just, let me just kind of mention as well to people listening, please put any questions in the chat. I should have mentioned that at the start and we're kind of more than halfway through now, but please, please do continue putting questions in and we'll, we'll continue to, to answer them. in real time or afterwards. Let's just move on to, or back to the kind of compliance side, if that's okay, and maybe compliance simplification. So how can AI, do you think, Simeon, kind of help simplify that compliance angle? Can you give us some examples around that? It's ever-changing regulatory environments that we work in, the requirements from there. So from an A to P SMS, how can AI help do that? If you can give us an example, that'd be great. Absolutely. Well, yes. So leading off with an example, I mean, we're seeing regulation come out that is common sense, very simple wording and incredibly hard to achieve. A great example of that is France. They came out with regulation about a year ago now saying that outside of sort of waking hours, people shouldn't be subjected to promotional messages. You know, it's almost like, if you like, a right to disconnect. So they said outside of waking hours, people can still receive one-time passwords. They can still receive notification messages, but they shouldn't receive marketing and advertising messages. So very simple to say, really hard to enforce if you're trying to do a traditional sort of keyword or, you know, regex based approach on traffic detection. You know, folks who are looking at this from the lens of a security perspective are now suddenly being required to understand the context of a message. And this is where machine learning can come in, because, yes, you can require your brands to. sign up back-to-back terms and conditions. But James, your earlier point, at the end of the day, the regulator is holding the operators who are then holding the aggregators accountable for that. There are levers and mechanisms in place to require people to conform. And this is a great example of where we're seeing machine learning as a perfect fit to be able to do this type of profiling of traffic to help educate. Because sometimes brands are well-intentioned, but maybe they're just not aware of all of these different terminating country regulations. And this is why I feel a combination of aggregators and campaign service providers utilizing this type of firewall can help not only regulatory compliancy, but just an overall better customer experience. Right. Anything to add to that, James? It's going to happen, Nick, on the basis of... The visibility needs to be there. But when you've got T-Mobile in the US on the 1st of January this year, introducing fines for specific fraud cases, that is going to force people to act. If you look at aggregators being over and above the fines, actually being removed from certain markets, they are going to act. If you look at your role in terms of meth and having worked with meth, the work that goes on around the codes of conduct. for markets, which where all the aggregators are very keen to sign up for, it wouldn't be great if we didn't need those codes of conduct. And actually, we just had visibility and facts, and the AI could actually enable a marketplace which we could trust in the future. That would restore the value chain that needs to, that is, it's going to happen through WhatsApp and RBM, but it would actually restore the faith in SMS. Right. So let's one final question from me just to kind of wrap things up and we'll get our crystal balls out now for a little bit. So it's always the interesting part. So, you know, getting your your views as the experts and I'll stick with you for now, James, if that's OK. But what what's your view in terms of future trends and what what do you foresee in the future for the ATP market? And how can maybe AI machine learning help providers stay ahead? AI is going to be used at the forefront of their customer experience strategy. First point. Second point, AI can help remove the manual processes that have been largely uninvested in. So effectively, the core operating model that sits underneath these, both of the M&O and the platform providers to improve the efficiency and enterprises will get access to that detailed information. And quite frankly, if they don't, they will act in the interim. They will shift and find a way of getting a trusted mechanism and delivery of their services. And I think to what you'd said before, in terms of T-Mobile US, it certainly focuses the mind when commercial penalties, I suppose, are coming into play. And you just naturally, to keep afloat or to keep yourself in the black. you have to comply with these things, it becomes mandatory. The reputation is far worse than any fine, the reputational damage. Yeah, we see on social media and other areas that if certain messages go out there, then it can have a really damaging impact, much further, broader than that specific issue. Simeon, if I can come to you on that question in terms of the future. Yeah, absolutely. I mean, I think... Well, echoing certainly James's earlier points as well, maintaining customers'trust. It's a fragile, hard-won thing, and it's so easy to lose that. And once it's lost, if customers aren't going back to that inbox app, if they're not going to their messages app, all of these discussions are for naught. So keeping that safe, keeping it clean, keeping it from the brands they trust is utterly critical there. I think in terms of… to the topic of, you know, how is AI going to help this? I mean, you mentioned it before, James, I think cost optimization, you know, businesses have been running cost optimization strategies based upon um blunt uh control points you know blunt levers where it's more you know what is the price on this day whereas what ai gives us is the ability to be surgical on that optimization you know what is the most appropriate control for this message and that may be different to the next message or the previous one so being able to do that as you say there james you know that that enhancement of business operating process but on a per message basis, unlocks a level of scale and business optimization that we've not been able to do before. And I think the other sort of crystal ball projection is we absolutely are all, I think, in agreement that rich media is the future in terms of some of the high levels of growth and delivering all of the things that we've been talking about on this call in a rich media environment. The only way to do that is also by leveraging AI. When we're talking image attachments, video attachments, audio, that is something that can't be controlled via simple text-based regexes or keyword detections. It will require the capabilities that we've established now using AI on text to then be extended to these rich media channels as well. Right. Thanks, Simeon. So I think we could carry on discussing this for double the time, at least, that we've been allotted. But I think we'll have to go back to that. to draw to a close there so thank you thank you both very much and i'll just leave it by saying that if uh if certainly you you uh are looking to boost your a2p sms revenues and even rich communications around ai and certainly seems like you need to speak to simeon and the team at ania and james myself as well so in terms of trust security safety clean messaging and business growth Hopefully we've got you covered. So thanks very much for listening and hope to see you again soon. Thanks.

So let's unlock the hidden potential of A2P SMS together. So I'll come to Simeon first. Hi, Simeon.

Can you please say hello and introduce yourself? Absolutely. So yeah, I'm Simeon Coney.

I am responsible for business development at Aenea. So I work a lot with our customers right across the globe, regulators and industry analysts, really looking at trends in the marketplace and how we can help with our solutions. Excellent.

Thank you very much, Simeon. And James, it's great to have you with us as well. Could you please introduce yourself? Yeah, thanks, Nick. So my name is James Lasbury and I work for Global Telco Consult, GTC.

We are experts in messaging and we help our customers do messaging better. But we also now work in mobile identity and other services. So effectively, we work predominantly with MNOs, so the operators globally, but also the biggest platform providers and enterprises across the markets. Prior to that, I worked for Telefonica and also Vodafone. and Vonage just after Telefonica, probably worth mentioning.

Excellent. So thank you, James. We've got a very similar background, except for the Vonage bit. So yeah, I'm sure we'll have a great discussion today. So let's start, Simeon, if that's okay with yourself.

So just, if you can just give us your general view on the state of the market for A2P SMS. Yeah, I think we're in interesting times, Nick, I really do. I think we've got a lot of external forces happening on the market as well as internal generated ones that are meaning we're probably heading for a very interesting time over the next 12, 18 months.

I mean, certainly we've seen the pricing changes that we commonly talk about having a very significant impact to the market. We've seen brands becoming more aware of not only the price points, but the value that they're receiving from messages in various markets and then potentially tailoring what type of communication they're doing based upon that price point. But also compounded with that, we've seen like a wave of, for example, behaviors like fraud through AIT occurring in the market as well. But yeah, not only that, the growth in rich media as well on some of those other bearers, I think it really means that we're going through quite a transitionary period right now.

And James, yourself? Yeah, I'd look at it slightly differently, as in being in the market for some time. I agree with you, Nick.

The market is in growth. That's a fact. However, if you break it down into the component parts, if we look at the biggest part of messaging today, which is SMS, that value chain is broken.

at the moment and trust is not there anymore. And what I mean by that is where it starts messaging from the consumer, from the enterprise to the MNO who delivers effectively that message into the consumer, that part in the middle where the aggregators and the platform providers sit, the trust across that ecosystem is really being stretched now. The impact of that is that some of the biggest brands in the market are choosing now not to send messages in specific markets due to the cost or even the quality of service.

And this is having a direct impact in terms of other channels. So the reason the market is growing, because if it was if it was just SMS, that is now plateauing and predicted to go down. WhatsApp.

And RBM, obviously RCS, is coming through now. And that is driving the growth. And the topic that we're covering today couldn't be more pertinent because a big part of the growth there is driven by bots and chatbots.

So I think this is a great conversation point. I look forward to discussing it. Right.

Yeah. So I think from what I'm hearing, the general trend is that the SMS is kind of peaked very much now. And in certain certain markets, it's declining. And I think, as you say, James, through the through the trust element is where we see most of the challenges right now.

But if I can stay with you for now, James, I mean, just, you know, in terms of A2P SMS, there's a lot of. challenges that we are navigating around enhanced traffic, abuse, increase in regulatory requirements. So, I mean, how are the providers tackling this or are they not?

Could you just give us your views on that? Yeah, it's an interesting topic, this, because every single provider in the market is very specific on how the brand should protect themselves. So that is stated within...

every service provider's website. And the reason that they state that and now provide tools for protecting their customers is because the regulator can implement, basically, they can enforce stops of service and huge fines within the M&O community, which are automatically passed on to the platform providers. So that's written within the contract.

And that now, over the last... probably two to three years, we're now seeing platform providers, because of smishing and some of the fraud issues, actually being facing high fines, but also being barred in terms of the services that they can provide into certain markets. So if we talk about kind of, is AIT new? How are the consumers being protected? That's always been there.

Are we facing a change in that? The answer to that is no, Nick. What we're facing a change to is how we start to deal with these issues, AIT and the high prices, because effectively brands are moving away.

And it's taken that for action to really take place in this market. And the way that we solve some of these problems are through kind of AI and machine learning, which is brilliant because platforms like... Simeon's that we obviously have at our disposal today are some of the ways in which we can block and deal with these issues. For example, I've worked with Simeon both within my past companies, Telefonica and Vonage, and had good partnerships in place to help kind of stop some of that abuse. And that's the key here, Nick.

If I can come back to some of the solutions in a second and the products. But Simeon, if I can... Just come to you.

I mean, in my position at METH, what I see a lot of is everything right now in the mobile world. And if there's two things that are, if you like, trending at the moment, one is RCS, which we're not here to talk about in great depth today. But the other area is definitely around fraud, anti-fraud, inflated traffic. But from what I'm hearing from all quarters of the industry or all...

divisions of the industries, that's rising and it's continuing to rise. So could you just give me your view on that? We'll talk about the products and the technologies in a second, but just give me your view on what's happening in the market, please.

Yeah, certainly. Well, I think the first thing is there are more points in that value chain who are getting better insights and understanding on what is actually being carried. I think absolutely, James, I agree with you that sort of... the what is being done hasn't significantly changed, but I think maybe we're now getting a better insight on the scale and the speed of change of various things within that, because we have better ways of sampling or inspecting or gaining control and understanding of what's happening in there.

So yes, I think what we're seeing now is that people are able to better quantify things like the different types of fraud. Things like the growth in AIT, I mean, certainly we are seeing not only changes in volumes, but very quick changes in techniques and behaviors. Again, you know, to the earlier point that there's a lot of external factors, you know, price points are changing so quickly in so many different markets, and the brands themselves are responding quickly to that, that we're seeing the frauds also being as dynamic as those two external forces.

So- You know, what was true three months ago is not true today. It will not be true of what's happening three months down the line there. People are capitalizing upon those dramatic changes to be able to either exploit for commercial value in terms of a competitive advantage or exploit from a fraud perspective to gain an advantage over maybe their other businesses. It's both positives and negatives, but you'll say that there's a great deal of change and the insights that people are gaining from those are helping them better understand what's actually happening there.

So if I understand correctly, there's the artificially inflated traffic that's currently happening, which is being driven by price rising. And again, please correct me if I'm wrong. I'm playing a bit devil's advocate here as well. But so my understanding is that the traffic's being inflated because the price is rising, because it can basically enable the fraudsters to get rich a bit quicker. Is that is that completely true or is there a bit more to it than that?

I think that's certainly one of the major factors. I would say another is the ability to programmatically generate large volumes of traffic, often through unsecured interfaces. What we've done in the ecosystem is enable brands and services to generate automated messages, but those services themselves may not be fully protected in the way that people can access and invoke them. So we're seeing a whole range of different mechanisms and techniques being used to generate AIT.

And often that does align to markets with high price points, but not exclusively. I mean, we are seeing it sweep across other territories where the price point is lower, in part just because there are brands operating in those markets who can be exploited and utilised to generate that AIT. Or in fact, people can use AIT not just for commercial financial gain of generating a message to get paid for, they can use it to generate an impact to their competitors'business. So things, for example, like conversion rate poisoning that we've spoken about within the meth community also has a consequential impact, even if it doesn't have a direct financial outpayment. It does surprise me ever so slightly.

I just, I... I was there at Orange when it launched. That was back in 94. And so it's probably giving you an indication of my age, which is a little bit scary. I don't know how it quite got to this point. But it does surprise me over 30 years, we've got this, if you like, fairly old now in technology terms, technology that's being exploited even more.

And it's becoming such a big thing in the industry, you'd have almost thought it would have gone the other way. So yeah. Again, I just find that quite an interesting scenario, really. But let's talk about kind of looking into the future a little bit and maybe talk about some of Aeneas'technologies and products that you have. And the buzzword obviously out there at the moment, if you look at any startups or any, I guess, subdivision of every business now, we're talking about AI and machine learning.

So could you just tell me a bit about... an ear's kind of role in that? You know, specifically, what products do you look at that tackles these types of things like fraud? Yeah, absolutely.

So, I mean, we've been using machine learning for over a decade now. And in fact, when we first started using it, we were very much using techniques that were looking for what we call lexical similarity. So if we think about, let's say, for example, a spam message, what was happening is spammers were saying, okay, well, This one campaign is blocked, so let's change word. You know, rather than you've won a prize, it's like you've been selected for a prize. So we were using machine learning techniques to catch those variations very, very quickly and also look at the mutation that happened within those campaigns.

So that was a system that worked really effectively for a long period of time to catch that type of investment that spammers were making to try and stay ahead of. sort of simple keyword-based and regex-based firewalls. What we're now doing is actually using the new wave of large language modeling to do not just lexical similarity but actually get to the meaning of the message with semantic similarity.

And this is where it gets really exciting because lexical similarity, I say, it's about looking at things like word sequence and word replacement, whereas semantic is very much about the meaning of the message. And with the large language models that we now have access to, it's given us the ability to abstract from the actual language being used. And again, we're talking here about the SMS marketplace.

It's a global market. And what's fascinating is the array of languages that are in active use by so many major brands. You know, people want to communicate in their language of choice.

So as a fraud protection security firewall, it's our job to be able to deliver that protection no matter what the language is. I mean, we today cover a quarter of the world, so we're encountering many languages. And when machine learning really comes to play, it enables us to have that abstraction of language and really get that understanding. So from not only a fraud perspective, we can handle now these campaigns that might jump from language to language where the fraudsters are using translation software.

But also we can use these same capabilities to gain an insight to the nature of communication to actually help in other not only value add areas, but also things like regulatory compliancy. where it's again about the principle of the message rather than the exact content that the regulators are getting concerned about. And I guess beyond the whole, I guess, area of fraud or anti-fraud, the applications for that type of, you know, view in the message language, the semantics, you know, has a whole raft of positive connotations where Customer care can manage their customers better potentially, or it could be a sales team could know if somebody is actually interested in buying the product that's being sold over a text message or over a mobile message. Is that what you've seen as well, James, I guess, beyond the whole fraud piece around, if you like, AI and machine learning? Yeah, so for people that don't know our space really well, and I know we all do, it's probably important.

to state that messaging, the reason that we have fraud and smishing is because messaging is so successful, right? It's been used in every market globally, and we'll see that evolve. And if we look at, say, for example, the strategy implemented at Telefonica Group across all of the operating businesses, so from Latam to Europe, it was very simple. It started with securing the market, then monetizing, and then... evolving the service.

And effectively, what we're talking about here is more effective security of the service with the products that Simeon has brought in. And if we look at it, kind of stretch that out to the MNOs and some of the other services, I know Aenea have been extremely successful in terms of obviously the number of global operators that it works with today. But we are still in a position whereby not every mobile operator has effective security tools. And if we take this back, Nick, to my first point around trust, unless we get the full visibility across the services that are being provided so that the MNO can make the right decisions, the enterprise doesn't have to ask for the tools. It has the visibility around what's going on with its customers and can act in real time.

That really then, that changes things because this isn't a question on whether messaging is a successful channel. It is. It's scaling. The issues that we face today are around bad actors and those generating fraudulent activities can't be tolerated in any way possible. Artificially inflated traffic, that's just to inflate revenues for a number of bad actors.

So in summary, Nick, I think you could almost flip this on his head. We face a really healthy future if we can solve some of these things. And it's brilliant hearing the way Simeon and his business is adapting.

But this isn't new. So I think Gartner predicted that 85% of enterprises would use AI in their customer communications in 2016. And that was their forecast actually by 2025. So that's a year away. It'd be quite nice study looking at whether that, how real that was separately, but effectively AI and machine learning is being embedded in all kinds of communications. And we can even look at some of those players in the value chain, for example, Twilio, Flex, a core part of the service is AI and machine learning.

If we look at GMS, one of the biggest providers, in messaging today globally for international messaging, they're effectively pivoting their business now towards AI to manage the customer journeys more effectively and really produce a brilliant customer experience. So I can see your dog just coming around your right shoulder, which is nice, Nick. Hopefully he or she can contribute a bit. I'll need to reposition my camera so you're not distracted. But at least she's not snoring anyway.

So that's a good thing on a live podcast. So would it be fair to say that some of this, you know, we're talking about AI and the products in there. But I'm. A large part of this is about visibility as well and opening up that visibility because you've got a number of players in the value chain, as it were, in the ecosystem, I suppose, that are potentially forwarding a single message. It hops through various different companies and through different routes.

Is visibility key to this or is it AI that's going to fix it? Do we need to show where things are going wrong? to everybody or should it be should it just be kind of nipped in the bud through this through the technology yeah it's it's an interesting point um simeon they if i give a context in terms of mno then you can give the context the security provider and and nick it's it you you know this as well messaging is incredibly successful but isn't invested in hugely by the mno it's seen as a channel that was growing 15 to 20 percent year on year largely by itself and therefore it didn't get the capex investment into the automated tools to manage the channel effectively so what you had is limited resource with largely manual tools which effectively we're talking about the wholesale shift in that which simeon's capabilities can deliver simeon i'll let you um yeah thanks i think you know what what ai enables us to do is on the visibility point, break apart the fact that it's just a message and go, okay, what is that message trying to do?

Who's it coming from truly? What are they trying to do? And with that, you can then start to align it to the value.

I mean, I think if we look at other channels, things like WhatsApp, to my mind, the great thing they've been able to do is do use case. based pricing. They've actually said, what is the nature of the engagement that a brand wants to have with their customer? And let's give the value of that conversation to the nature of that discussion. So it's not priced per message.

What we want to do is get an outcome. We want to get somebody who's activating a service. We want to have someone who's getting the appropriate level of support.

We want to have somebody notified about a service interaction. And I think this is where we see, again, machine learning. helping and gain an understanding of exactly what the transaction is and maybe moving the market away from this sort of commodity unit based price, which only varies destination operator, destination country by country into something that is far more aligned to the value that a brand and a customer is actually trying to achieve. So on that note, in terms, because I would expect the majority of people listening to this webinar are going to try to understand one or two things to understand this industry a bit better and understand our views on this.

But the second thing is, and I hear this all the time is, what opportunities are there to monetize these types of technologies? So we talked about what it kind of almost does to kind of try and stop the bad actors and the tools that you have in place to be able to do that. But on the flip of that, how could we use this type of tech to to help enhance the monetization.

If you've got an answer on that, Samir? Yeah, absolutely. So that's the key thing. A firewall isn't an inspection and a control point.

The fact is that it's been used for fraud and security to start with. But to find new fraud, you have to gain an understanding. To work out what something bad is, you have to understand the context of that and you have to know what good looks like.

So yeah, a firewall is an ideal control point. to be able to gain these understandings to say, okay, what is the nature of the transaction? I mean, James, you were talking about regulation as well earlier. This is where firewalls are no longer security points, but they are also regulatory and compliancy points. But off that same back, if you're doing those two functions, why not now start to leverage it for these other value-add services?

I mean, I mentioned WhatsApp. I mean, a great example is now starting to profile traffic in our customers'networks and then doing a comparative analysis to WhatsApp. It's really interesting looking at the breakdown of, you know, what's notification, what's promotional traffic, what's one-time passwords. and then applying a WhatsApp pricing model to do the what-if question of, is WhatsApp cheaper or more expensive than SMS? You know, when you're just doing a unit-based price, it's really hard to try and work out that apples-to-apples comparison.

But with AI, you get that ability to profile the traffic. And I'll answer the question now. In a number of markets, SMS is actually priced below the price point that WhatsApp is achieving. So this is where we see there's a real future in using AI to gain this profile, to maybe evolve pricing models and again, start to align traffic to the value.

And to me, it seems like, I mean, ultimately it's the brands or the enterprises or the businesses out there that are communicating with their end consumers that are paying for this. And my experiences have been they don't, especially if you're talking to, say, a procurement team, they don't get that. they just see the unit price and is that your experience as well because certainly it's been it's been mine that you you're almost positioned a price per message sms versus price per message on some other some other channel like rcs or whatsapp and it's really only until you can demonstrate maybe the roi on that so if you're selling i don't know if you're a hardware store and you're selling a chainsaw or something can you you you're increasing the conversion rate you two times through one channel versus the other.

It's only at that point then that you can sort of demonstrate the cost comparison. I mean, have you experienced that as well? And does AI help with that?

James, I think I'd look probably more to you as a change support expert. Cool. So here's the thing, Nick.

You'll know on the operator side, the cost base is fixed. So the cost is known. And I would challenge your view to it or your point to a certain perspective from is it the enterprise that doesn't understand?

Or is it the effectively the partner that's been conditioned into understanding the cost base, not actually communicating with the customer in terms of the value that can be delivered? and really understanding and helping the customer understand the difference between an authentication to a marketing message to something that could be a kind of conversation in the future. From my perspective, managing kind of the carrier side of a platform provider, where the job was to effectively manage the costs throughout the business on an ongoing basis, it was from a routing perspective, Nick. That was such a huge part. If you look at it from a platform provider, it's almost 80% of their cost when it comes to ATP, right?

So when flipping that into a customer communication, when the cost is such a high proportion of the actual kind of, of the service that's been provided, the way that the services have been provided today is almost on a trading basis. So they'll deal with a customer looking across a number of markets. and they'll aggregate the price point to give them something more effective. What we're now seeing and what the beauty of AI is actually communication solutions coming out.

We touched upon GMS. Their strapline in terms of creating a brighter future for their customers is all about using AI. Twilio has always been at the forefront of this area. They've always had kind of developer tools, but very much. in terms of solutions to the customer.

It's really easy because America was their biggest market and then they've expanded outside of that. The other aggregators that have served the other markets have been predominantly focused on how do they control the costs of their service across multiple markets. And I think that's a really important point to understand because sometimes on the M&O side, you don't see that.

Sometimes on the enterprise side, you don't get that. It's almost all of that information. is with the kind of aggregator or the platform provider. And this is where another area, Simeon's tools can really help in terms of the effective routing of the messages globally and the transparency of that.

Imagine Nika in a world whereby we wouldn't have to talk about hops. You just knew who was delivering the message and then you could see the conversion and then you could decide based upon that who you used. um thereafter i mean we're we're talking about um some pretty you know, groundbreaking stuff, I think here.

And wouldn't it be great if, as a brand, you could almost, you know, in real time, see that the cost of a message or something that's being delivered securely and the benefit that that's actually giving you. And just maybe a silly example of that is, I mean, majority of SMS today is one-time passwords from what I see, certainly A to P. But that has a value associated with it. So if...

you know, the mechanism through a text message isn't slick enough so that that person is logging into the website. And then let's say you deliver that same service OTP through another channel or some other mechanism that just makes it easier, makes that person maybe log on more, which then has a knock-on effect to drive in customer engagement, customer loyalty, or even some sale of some description through a retail website, then wouldn't it be great if you could demonstrate that through the power of AI. Personally, that's the sort of thing if I was heading up a customer care or an IT team or I was the CMO of a company, that's the sort of thing I'd probably be looking for. So again, I don't think we're too far away from that, from what we're discussing here. Nick, I can just jump in on that point.

So to give you a real data point in terms of how AI is helping our insights in the industry. What we're now seeing is one-time password is no longer the dominant traffic. I mean, we know that it was going to be affected by the increase in price points and the rise of things like authenticator alternatives.

But using that traffic profiling techniques, we're now seeing that actually notification is now the dominant traffic type in A2P flows. One-time password still sitting at number two and then promotional traffic actually showing the greatest level of growth. in the segment there.

So it sounds like brands are starting to think about how can I use promotional messages to sort of re-engage my customers. And as we all expected, the one-time password traffic starting to drop, but actually notification, not only by traffic volume, but also by the number of campaigns, the number of distinct campaigns also is the largest segment now with the highest growth. So again, I think that points to a very, very positive future and the outlook of... how brands are using messaging.

Yeah, that's really interesting. Again, that's probably my, you know, naivety at the moment around SMS that, you know, most of the conversation seems to be around the one-time password. But if indeed it is notifications, it is promotional messages, then certainly with the power of AI and other maybe richer channels that are coming to market, you know, we should see a lot more benefits of using those.

And I think Stats that I've seen from a couple of different research companies, and I'm sure you know the names of them, which I won't mention here, but they're all saying that the number of enterprises, the number of brands is very low compared to the number of brands that could be using these types of technologies. So, you know, if the messaging providers can really expand the portfolio out to more businesses and we get more usage of these types of technologies, then, you know, I think. again, bringing AI into the conversation and maybe even rich messaging, you know, you become even more successful with this, with this area.

So let me just, let me just kind of mention as well to people listening, please put any questions in the chat. I should have mentioned that at the start and we're kind of more than halfway through now, but please, please do continue putting questions in and we'll, we'll continue to, to answer them. in real time or afterwards.

Let's just move on to, or back to the kind of compliance side, if that's okay, and maybe compliance simplification. So how can AI, do you think, Simeon, kind of help simplify that compliance angle? Can you give us some examples around that? It's ever-changing regulatory environments that we work in, the requirements from there. So from an A to P SMS, how can AI help do that?

If you can give us an example, that'd be great. Absolutely. Well, yes. So leading off with an example, I mean, we're seeing regulation come out that is common sense, very simple wording and incredibly hard to achieve.

A great example of that is France. They came out with regulation about a year ago now saying that outside of sort of waking hours, people shouldn't be subjected to promotional messages. You know, it's almost like, if you like, a right to disconnect.

So they said outside of waking hours, people can still receive one-time passwords. They can still receive notification messages, but they shouldn't receive marketing and advertising messages. So very simple to say, really hard to enforce if you're trying to do a traditional sort of keyword or, you know, regex based approach on traffic detection. You know, folks who are looking at this from the lens of a security perspective are now suddenly being required to understand the context of a message. And this is where machine learning can come in, because, yes, you can require your brands to.

sign up back-to-back terms and conditions. But James, your earlier point, at the end of the day, the regulator is holding the operators who are then holding the aggregators accountable for that. There are levers and mechanisms in place to require people to conform.

And this is a great example of where we're seeing machine learning as a perfect fit to be able to do this type of profiling of traffic to help educate. Because sometimes brands are well-intentioned, but maybe they're just not aware of all of these different terminating country regulations. And this is why I feel a combination of aggregators and campaign service providers utilizing this type of firewall can help not only regulatory compliancy, but just an overall better customer experience. Right.

Anything to add to that, James? It's going to happen, Nick, on the basis of... The visibility needs to be there. But when you've got T-Mobile in the US on the 1st of January this year, introducing fines for specific fraud cases, that is going to force people to act. If you look at aggregators being over and above the fines, actually being removed from certain markets, they are going to act.

If you look at your role in terms of meth and having worked with meth, the work that goes on around the codes of conduct. for markets, which where all the aggregators are very keen to sign up for, it wouldn't be great if we didn't need those codes of conduct. And actually, we just had visibility and facts, and the AI could actually enable a marketplace which we could trust in the future.

That would restore the value chain that needs to, that is, it's going to happen through WhatsApp and RBM, but it would actually restore the faith in SMS. Right. So let's one final question from me just to kind of wrap things up and we'll get our crystal balls out now for a little bit. So it's always the interesting part. So, you know, getting your your views as the experts and I'll stick with you for now, James, if that's OK.

But what what's your view in terms of future trends and what what do you foresee in the future for the ATP market? And how can maybe AI machine learning help providers stay ahead? AI is going to be used at the forefront of their customer experience strategy.

First point. Second point, AI can help remove the manual processes that have been largely uninvested in. So effectively, the core operating model that sits underneath these, both of the M&O and the platform providers to improve the efficiency and enterprises will get access to that detailed information. And quite frankly, if they don't, they will act in the interim.

They will shift and find a way of getting a trusted mechanism and delivery of their services. And I think to what you'd said before, in terms of T-Mobile US, it certainly focuses the mind when commercial penalties, I suppose, are coming into play. And you just naturally, to keep afloat or to keep yourself in the black. you have to comply with these things, it becomes mandatory. The reputation is far worse than any fine, the reputational damage.

Yeah, we see on social media and other areas that if certain messages go out there, then it can have a really damaging impact, much further, broader than that specific issue. Simeon, if I can come to you on that question in terms of the future. Yeah, absolutely.

I mean, I think... Well, echoing certainly James's earlier points as well, maintaining customers'trust. It's a fragile, hard-won thing, and it's so easy to lose that. And once it's lost, if customers aren't going back to that inbox app, if they're not going to their messages app, all of these discussions are for naught. So keeping that safe, keeping it clean, keeping it from the brands they trust is utterly critical there.

I think in terms of… to the topic of, you know, how is AI going to help this? I mean, you mentioned it before, James, I think cost optimization, you know, businesses have been running cost optimization strategies based upon um blunt uh control points you know blunt levers where it's more you know what is the price on this day whereas what ai gives us is the ability to be surgical on that optimization you know what is the most appropriate control for this message and that may be different to the next message or the previous one so being able to do that as you say there james you know that that enhancement of business operating process but on a per message basis, unlocks a level of scale and business optimization that we've not been able to do before. And I think the other sort of crystal ball projection is we absolutely are all, I think, in agreement that rich media is the future in terms of some of the high levels of growth and delivering all of the things that we've been talking about on this call in a rich media environment.

The only way to do that is also by leveraging AI. When we're talking image attachments, video attachments, audio, that is something that can't be controlled via simple text-based regexes or keyword detections. It will require the capabilities that we've established now using AI on text to then be extended to these rich media channels as well.

Right. Thanks, Simeon. So I think we could carry on discussing this for double the time, at least, that we've been allotted.

But I think we'll have to go back to that. to draw to a close there so thank you thank you both very much and i'll just leave it by saying that if uh if certainly you you uh are looking to boost your a2p sms revenues and even rich communications around ai and certainly seems like you need to speak to simeon and the team at ania and james myself as well so in terms of trust security safety clean messaging and business growth Hopefully we've got you covered. So thanks very much for listening and hope to see you again soon. Thanks.

Transcript for:AI's Impact on A2P SMS Industry

Transcript for:
AI's Impact on A2P SMS Industry