Twilio Inc. (TWLO) Earnings Call Transcript
May 21, 2024
Earnings Call Speaker Segments
Good morning, everyone. I'm Mark Murphy, Software Analyst of JPMorgan. Great to see everyone. And it is a real pleasure to be here with Khozema Shipchandler, who is the CEO of Twilio. So first off, Khozema, I want to thank you for taking the time, making the trip out here, and welcome to the conference.
Thanks a lot, Mark. It's my pleasure.
I want to also congratulate you on your new role this year. And hoping we could kind of kick off here by having you give us just a brief introduction of the company for the benefit of anyone in the audience who's not familiar yet.
Yes. Again, thank you. It's been a super exciting time, super exciting time for the company. Twilio, at its core, is a customer engagement platform. We enable that through core communications technology, a data platform and increasingly, AI, which I'm sure we'll talk about during the course of the conversation. But ultimately, our job is to make sure that we deliver really important workloads from our customers to consumers like you and I, using communications as a vehicle to do that.
So there has been an ongoing evolution to Twilio. There's been, frankly, a lot of change in the last couple of years across the industry and in the company, right? You've had changes that flow across the product. It's across the leadership, the go-to-market, the head count and I think importantly, the profitability structure of the company. How do you think that those changes might reinvigorate and perhaps set the business up for success if we're looking forward a couple of few years down the road?
Yes, it's a great question, Mark. I think from a change perspective, I'd maybe kind of call out 3, and I think these are what lead to the setup in terms of the medium- to long-term picture. I think first, we went down the path of creating business units for the company. I think that's led to a much better focus in terms of the way that we're allocating capital, the way that we're allocating our resources, and it's also allowed us to serve customers better because the go-to-market motions from the businesses are just kind of fundamentally different. And so that really helps. I think the second thing is that while some of the management team is the same, some of it is different, which I'll talk about in a moment, we're running the company totally differently than we were just 6 months ago even. And what I mean by that is we're running it with a lot more rigor discipline, focus, and I think that's leading to some of the financial outcomes that you're seeing, especially around profitability and cash flow. And this team is very, very serious about making sure that we meet those financial commitments. And then I think the third thing is, is that we've also been fortunate to be able to bring in some new blood too. So for example, we brought in Thomas Wyatt, who's leading our Segment business, Inbal Shani who leads R&D for our communications business. And then most recently, Chris Koehler who leads our marketing function. And I think the combination of some of the existing players as well as the new blood and in addition to some of the free cash flow generation allows us to start thinking about a new investment cycle, one that's done with a lot more focus than I would suggest it was done previously. But that's kind of the setup that is going to lead ultimately to medium to long-term growth.
And the cash flow has been a good support for the buyback activity as well, using the balance sheet and the cash flow. We went back and looked at this. You bought back stock 5 quarters in a row. I believe you're trying to exhaust the $3 billion authorization by the end of the year, right? So it's a pretty substantial reduction of the share count. Why is Twilio buying back so much stock? Especially, at a time when the cash earns a good return, just sitting there on the balance sheet. I think from an external viewpoint, presumably, you see some disconnects, right? Where the market might be kind of underappreciating, where Twilio could be, while you're kind of doing the buyback. Can you walk us through the mindset on this and what do you think some of the disconnects might be?
Yes. I mean I'd say fundamentally, the reason that we're doing the buyback is because we believe that the stock is undervalued. And so this is a really smart and opportunistic time from our perspective. You pointed to some of the cash that we have on the balance sheet. I think given the fact that we're in a free cash flow position, we're not really a significant participant from an M&A perspective. I think that gives us significant optionality to be able to buy back as much stock as we have. You're right on the numbers. We'd previously authorized about $1 billion, now $2 billion. We're aiming to complete the whole thing by the end of the year, and I think that positions us really well. In addition to that, we're starting to generate our own free cash flow, as we talked about. I think that also leads to kind of further optionality. We could buy back more shares at some point down the road. We could potentially look at M&A if something became attractive. But I think sitting here right now, the bottom line is, is that we do think the company is undervalued. We do think we've got a lot of optionality. And so being in a position like this is really fortunate and we'll see how things shake out after the end of the year.
So Khozema, let's look at the communications side of the business. About 93% of revenue, you're approaching this $4 billion run rate on that side of the business. And thinking back to the time of the IPO, Twilio very clearly has been a leader in this market, probably close to a decade ago, really define how this -- redefine how the market works. What inning do you think we're in at this point if we just were to look at SMS, text messaging? Because I think on the upside, we look at it and we say, this is a market that has ubiquity, you have the high open rates, you have reliability. There are also alternatives, right? And people will bring up push notifications. They will bring up -- there are other messaging constructs like WhatsApp and Facebook Messenger. How do you think through this market and try to understand whether there's a little bit of maturation going on or whether there's still a lot of runway left?
I mean I think it's mostly a tailwind. And I'll describe it this way. I think, first of all, we're fortunate in that we serve all of the different channels that you just mentioned, okay? So whether or not communications end up transiting through OTT, through SMS, whether -- I personally think there's like sort of a renaissance that's happening in voice right now, especially with AI workloads. And it's just like so much more natural to talk like you and I are versus doing it necessarily through some of the other channels. So I think that's going to be kind of an accelerant too. But I think all of those channels ultimately value ends up accruing to Twilio because we support all of them. I think each of them ends up serving different kinds of workloads. I think when -- just, let's use push as an example or kind of in-app messages, like the reality is, is that for something like that, most of us have tolerance for a certain amount of real estate on our phones, for example. And so we're only going to take so many of those things and which is what makes SMS so attractive. It's the universal standard. It's hard for me to imagine sitting here today that something supplants it. I'm not a huge baseball fan. I realize it's probably sacrilegenicity, but well -- it's also the MLB playoffs, I guess. So let's say, like it's probably the second period in terms of where we are with respect to that. But I think there's a lot more that's going to happen here. And we're seeing stabilization in volumes right now, but I think especially based on AI and what we're able to do there, we do think there's an opportunity for some additional tailwinds as things go.
Okay. So a second -- second period of the basketball game for SMS type, but you're hopeful or optimistic on some tailwinds?
I'm kind of both. I mean, I think we've given guidance that we're not going to veer from, right? So we feel good about the kind of guidance framework that we've got. We haven't put it out there for longer than the period that we've given. I think medium to long term, we feel quite optimistic about the direction of the business. I think whether that ends up accruing exactly to SMS or through some other kind of communications channel, I mentioned voice, I think, is going to see a renaissance over a period of time, I think that remains to be seen. But generally speaking, I'm extremely optimistic about the prospects of the business.
Okay. We may need to take another look at what's happening on the voice side of the business, and I appreciate you mentioning that. So Khozema, I'm thinking back to 1 year ago, it was on this stage that Twilio had mentioned kind of starting to see a stabilization of usage volumes across the communications business. So that would have been Q2 of 2023. And people in the audience certainly noticed that, right? At that time. Thinking back to -- you had your recent earnings call about 2 weeks ago. And there was a comment that messaging volumes continue to stabilize, but they weren't. There was a comment that they're not exactly inflecting upwards. And you said -- also said on the earnings call, we're focused on growth levers we believe will drive reacceleration both -- in both the short term and the long term. So I think we're curious what it is that you're alluding to. And do you see some kind of uplift potentially coming in that organic communications growth rate? At some point above because that level was reasonably good. It was 7% level there in Q1.
Yes. So a lot there. So let me just kind of unpack it one at a time. I think, Mark, in terms of the stabilization, like we continue to see stabilization in volumes. But as we noted during the time of the call, what we've not yet seen is that inflection up. And so I still think we're kind of in that zone, and we're optimistic about that potentially transpiring. I'll talk about that through some of the efforts that we're undertaking ourselves to drive our own growth. But I think in general, we see a stable picture. We don't see sort of an inflective picture if that makes sense. I think in terms of some of our own initiatives, in particular, the ones that we're really focused on are cross-selling our products in different parts of the business. Once we have a landed customer, it's obviously a lot easier to be able to continue growing at that customer through other different products and services. And I think we've seen examples of that. In particular, we've seen good examples with Fraud Guard and Verify Fraud Guard, for example. We've seen it with Voice Intelligence. I'm sure we'll talk about that later. But I think those are good examples where they represent upsells ultimately, they represent higher prices, better margins. And so we like the characteristics that cross-sell ends up driving I think to -- we've started to do our own work with respect to segment and the communications business, bringing those 2 capabilities together. I think that will drive its own kind of cross-sell. And there's a specific product that we launched around that. And then finally, I think the other area where we're seeing a lot of traction is with respect to ISVs. So ISV volumes have been very, very strong. I think ISVs, in particular, allow us to reach a much broader set of customers and then ultimately, consumers. And so that growth has been pretty exciting as well. I'd say all of those are things that we're kind of doing in the short term. I think in the medium term, ultimately, it's going to be, I'd say, predominantly, if not all, organic, like we're not really planning for any inorganic activity right now. So the organic activities would be predominantly behind investments that we're making today that we're self-funding as a result of our own profitability and cash flow that I'd suggest probably start to show up next year, late next year, in subsequent years, et cetera.
And I want to come back to the Fraud Guard product in a couple of minutes and also some of the ISV wins that you had announced. Before we do that, can you comment a bit on the ADP and traffic registrations. Are we at a point where that is no longer a topic or are we through it? Because that had constrained some of the volumes for a period of time. I think actually, that was an aspect where the headwind was not as severe as you had maybe thought it would be in the middle of last year. Are we through that?
Yes. The short answer is yes, and it didn't really have a significant impact at all in our business.
And how about the impact of this movement over to the self-serve motion? Is that -- are we at a point where you feel like that is something that is kicking in for the Comms business where you can -- I think the main part of the thrust there was to try to onboard more business with less of a human touch. I don't know if some of that related to the support organization as well.
Yes, that's an astute question. So I think maybe just to kind of go back in time for a moment. So when we were kind of -- maybe a little bit over a year ago, last year, that was sort of on the back of a couple of layoffs that we'd undertaken. And at that time, what we felt, and I think we're now starting to see it bear out, is that we could be a lot more effective about reaching our customers. And so part of the layoffs were aimed at being more efficient our go-to-market motion and transitioning a lot more of that workload over towards self-serve. And in the most recent quarter, I think what you've seen is a pretty significant increase in our overall customer count. A lot of that ends up coming through self-serve. The reason that's so exciting is that some of the today's best-known brands, especially among the digital natives, they started to self-serve customers without us even knowing that they were going to end up being some of the world's largest brands. And so that's been great. I think our efforts, in particular, were really aimed at cleaning up a lot of the -- kind of keeping the lights on style work, compliance work, in particular, that needed to be done so that the friction for a self-serve customer was taken out of the process that ADP introduced a lot of that. I think we're past that now. So I feel quite good about where self-serve is. I think it's always been an important growth vector for the company. It's a fast-growing part of the business, and I'm really excited about going forward.
Khozema, how are you reading the tea leaves on the macro demand environment? Because I think what you said going back to what you said again a moment ago, it is -- there's some stabilization. There's nothing really kind of conveying a V-shaped recovery. I think that's certainly the sense we get from talking to other companies here. But I'm wondering what your sense is of how it could trend going forward the next several quarters? And I try to think from the seat that you sit in, are there -- are there kind of metrics under dashboard that you can watch? Is it -- are you trying to watch the messaging volumes in real time? Is that your best read? Or is there a way you could hone in on something specific, like looking at what's maybe the sense around -- the sense in recent weeks and months is that -- it's a small business that's struggling more? Small business is struggling with the weight of higher interest rates. Do you have any way of teasing it out in your data or anything in particular that you look at?
Yes. I wouldn't characterize our business exactly in the way that you said, Mark. So I wouldn't necessarily say that we're seeing a read-through based on the volumes in our business where I would specifically point to like an SMB style impact. In fact, if anything, I think just given what we talked about a moment ago around self-serve, like a lot of those businesses tend to be fundamentally SMBs, even if they're digital natives. And so I think that volume has been actually pretty good. I think generally speaking, we're seeing most industries have been generally pretty solid. Like again, no one is inflecting, I think, in a necessarily like really standout way, but I think as I look at like the demand signals in the business, and I think what you're really getting at is like, our business is sort of a leading indicator in terms of a lot of industries. We are seeing strength in a variety of pockets. We're seeing it. Honestly, in most industries, there's one exception to that, which is kind of international termination. But broadly speaking, we see a pretty good environment. We're not planning for in our own FP&A planning like a macro recovery of some sort. It kind of seems like it's going to play through. I'm not an economist, so I'm not really going to take a view on that, one way or the other. We're not planning for a lower interest rate environment, like that's not what's like really informing our models. I think if it happens, it probably provides a tailwind, but that's not really how we're planning for our business. I think based on the demand signals that we see, based on the workloads that we end up carrying, it's been generally pretty solid.
Are you able to get a read on how your larger customers are thinking about their marketing budgets? Because there's been -- we have been through this period where when you get into an economic slowdown, right? They tend to pull back to some extent on the marketing side of it. I'm wondering if you're able to get a read on that? Or even just looking at retail, e-commerce related messaging volumes, I don't know if you're able to see what is in order of confirmation or like anything that would kind of inform you the health of the consumer?
Yes. There's maybe 2 different dynamics there, actually, Mark. I think maybe just one thing to be clear, look, we definitely do not look at the content of our messages. So like that would sort of violate our own use policy with customers. But in terms of the industries and then the workloads, obviously, we know kind of the kind of messages. I'd say retail has actually been pretty strong, retail and e-commerce. Like it's not been crazy strong. But I think, generally speaking, I think from our perspective, the read-through is that consumer spending as it relates to retail e-commerce has been pretty decent. Obviously, I understand that based on the number of companies that have reported and the results, there seems to be kind of like a mix, maybe even soft picture out there. But I think in our own business, the volumes and workloads that we've been carrying it would suggest that retail e-commerce has been okay.
Okay. How do you think about staffing the business appropriately, kind of for the environment that we're in? Because we do keep harping on this, it's been across the software industry, the hiring has been very sluggish. And then when we look at Twilio, your head count compressed, I think, for the sixth consecutive quarter. You are one of the businesses that adjusted very and tried to right size very rapidly. But is there anything that would inform your view that it is at the right time to kind of lean in again on hiring? Or do you think we look at something structurally where you -- because of the nature of the business, you could see a reacceleration when it comes around -- without really having to incrementally ramp hiring?
I don't think those 2 things are necessarily connected. So I think in our -- if you're suggesting, is there revenue growth, potentially reacceleration or what have you tied to head count growth? I would say not for our business necessarily. I think our staffing levels are quite appropriate. I have no regrets whatsoever about the actions that we undertook over the course of the last year. So I don't think it's really impacted growth in any significant way. On the contrary, I think it's led to much better process and communication and efficiency inside of the company. I think it's delayered. I think that makes things just kind of simpler. It takes friction out of the system. And I think the way that we're planning today is that we don't necessarily need any additional head count to be able to drive additional growth. I think one other thing I would just add to that, Mark, is that, especially now given some of the advent of like what AI is doing for a lot of companies in terms of their external workloads and us, too, I mean, the same things are starting to apply internally as well. And so I think, if anything we see opportunity as it relates to AI, to automation, and I think that actually helps us with operating leverage over time in addition to kind of the natural lift that we'll get from volume leverage.
Okay. Okay. So in other words, yes, the top line growth can be pretty decoupled from the head count and staffing level. So what about -- as we think, Khozema, in terms of how it might translate into free cash flow, this is one -- heading into Q1, we had shared our view that the free cash flow generation over time could be quite a bit higher than investors might expect. And coming off of that quarter, you guided free cash flow 35% above consensus. I don't think we're going to see that too often. And so it's trending towards $4 a share for this year, it would have been around $2 a share last year. Do you have any perspective on how to trend it beyond that? In other words, so we've had the big surge. Is there some ongoing margin expansion while you've got a buyback happening there in the background, right? That, to me, is a potential equation on being pretty supportive of free cash flow per share.
Yes. I mean again, there's a lot there, Mark. I think the free cash flow generation, put simply, is like really on the back of our operating profit, right? I mean I think the business has started to deliver really significant profitability. I think that in the first half, especially of last year, we were undertaking restructuring, right? And so that has sort of the cash outflow impact. I think as you normalize out of those different things, your cash flow tends to kind of track operating profit. We're not like a significant working capital user. It's not like we have like significant CapEx in our business. And so by and large, we feel pretty good about free cash flow's ability to kind of track where we are from an operating profit perspective. What we've said, as a part of our guidance, is that we expect free cash flow to be in line with our operating profit and operating leverage. I think that over time, I would anticipate that as we deliver additional operating leverage, you should expect kind of the same thing from a free cash flow perspective, given what I said about the balance sheet characteristics of the company. I think that the buyback and the free cash flow, I mean, they're coincidental, I would say, for the most part. I mean, it's not like we did the buyback because we necessarily saw so much free cash flow or necessarily saw urgency about generating so much free cash flow because of the buyback. I think that they're coincidental events. We feel really good about the ability to do both at the same time. Obviously, doing that significant of a buyback and taking down the share count while also generating this much cash gives us a lot of optionality, right? And so I think that could allow for additional buyback down the line, that could allow for M&A if we felt like there was a good fit. The bottom line is there's optionality there, and we're really proud of what we've been able to accomplish in a relatively short period of time, to be frank.
Functionality is a good thing. So Khozema, I want to ask you just for a moment on product strategy and one of the directions that I think you've been laying out with us, which is -- we think of this as cross-pollination of segment, the segment or AI into the comms platform, right? And so I think we've always appreciated that there should be some potential that you would take a CDP like segment. And you can then take customer data, you can plug that into an LLM or some type of an AI model, right? And then you're taking action via the comms platform. So we've always appreciated there's some kind of potential there. And going back to the recent earnings call, again, you talked about unified profiles. You also talked about Agent Copilot. There's a lot going on with that. And I think you're seeing in beta testing, you're deflecting 70% of support cases with some of these products. Can you make the concept there a little more tangible for us, maybe just in terms of what's the pain point that you're trying to address here?
Yes, definitely. So look, first of all, I think this really kind of goes to like disciplined focus and rigor. Like we've long said that these 2 product capabilities are better together and we hadn't delivered against that, right? So shame on us. Now we're very, very like deadly set on ensuring that we deliver 3 products this year where Segment is natively embedded inside of our Communications capabilities. The first one being what you referred to, which is Agent Copilot. Now in terms of what it does, okay? So the unified profile is basically a data view, contextual data view of your customer, and the Agent Copilot is AI, leveraging our contact center capabilities where we're able to determine during the actual context of a conversation using information that we know about the individual that's calling in how exactly to solve that workload faster. And in the example that you cited, it was actually done with an Online University based out of Mexico. They were running extraordinarily high CSAT rates, kind of mid-90s. They asked us to experiment with this with them. They wanted to reduce their costs. Their costs came down 30%. There -- the deflection statistic that you pointed to was 70% in terms of call deflection to some other kind of form of self-serve or whatnot. And in the process of doing so, CSAT stayed exactly where it was, right? So that's kind of like the promise of AI, right? Like it's a fundamentally same or better experience at a lower cost point. And I think that what we are -- the unified profile, you're going to see increasingly emerge in all of the capabilities that we end up delivering at Twilio that have Segment natively embedded and then these different Copilots whether we call them or not, in all cases, I don't know. But in this example, it's able to take that workload and accelerate what would have otherwise happened. Really pretty cool, actually.
Do you feel like it's something that's pretty packageable and replicable?
Yes. I mean I definitely do. I mean, I think that we've done our own experimentation with a lot of customers. And fundamentally, when a customer sees what's possible, I mean, their eyes light up, right? Like everybody runs like a customer service center or a contact center or a call center, whatever the case may be, and you're looking at these examples. And the takeaway from a customer's perspective always is, wow, I'm at least getting the same thing and I'm getting it for a lot cheaper. So I think that's really, really compelling for customers, especially if we are in a constrained macro environment.
So I think in and around that discussion, Twilio Voice Intelligence is something else that came up, right? It's handling sentiment analysis, it is doing call summarization. And along those lines of this product, you mentioned hundreds of customers are in the beta test. What are you seeing there? Is that something that's going to be pretty monetizable?
Yes, very monetizable. Again, it's an upsell relative to what we would otherwise ordinarily do in voice, which is already a pretty attractive product. As I mentioned earlier, like in the same way that you and I are talking right now, it's just like a much more natural way to interact. Like imagine if this was just entirely happening through text, right? Like that would be kind of difficult. You wouldn't get any sense of like sentiment, you wouldn't get any sense of like emotion and stuff like that, right? And so in Voice, what's interesting about the Voice Intelligence product is if a call is going totally haywire, like the AI can automatically detect that and auto escalate it, right, to a different person. If there are keywords that are super essential to, let's say, a retail transaction, right? Like all of that information can be harvested, it can be stored. And so that when the next interaction takes place between a consumer and one of our customers, we can make it better. And we can actually make it feel like you've been a customer of that business previously and that they know a little bit about you. And these were all like sort of dreams before in some respects, like they're now all possible. And ironically, like if you just kind of analyze it through the lens of like your own life, for my own life, these experiences are extraordinarily rare. Like we just like don't get them that much, right? And so we see the promise being really, really significant because each of us as consumers wants this so badly.
And then -- so another one, Khozema. I think you had brought this up in the first couple of minutes, it is Verify Fraud Guard. I mean, another example of kind of embedding AI into another product. In this case, that's the verification product. And we went in and looked at this, it looks like it's trying to identify unusual behaviors, maybe someone is creating multiple accounts, using the same phone number. Maybe there's something else going on. How do you think about monetizing and packaging that one up? Is it going to be something where someone today is using your basic -- your basic multifactor authentication, right, your basic -- or even voice authentication and that you're going to have an upsell off of that?
Yes. It's predominantly related to what happens around SMS. I mean, you can do it for voice, too, but it's predominantly geared towards SMS. So just to back up a step, like there is a lot of fraud actually that takes place vis-a-vis SMS. And we strongly recommend each, every single one of our customers that, at a minimum, they take kind of the free version of Fraud Guard that we offer them. And if they take a paid version, we actually guarantee it at 100%. And as it relates to Verify, it's an incredibly fast-growing product. It is an upsell. And the reason it's so attractive to a customer is it pays back in as fast as 30 days, right? So the ROI is pretty incredible. And oh, by the way, like we actually send every one of our paying customers an e-mail every single month that shows them exactly how much fraud was stopped, how much money that they saved as a result, and that makes it a very sticky product as well. And I think just more broadly, like we all want like trust in the ecosystem, like these channels are more valuable as long as they're protected and products like Verify Fraud Guard would allow for that.
Where do you stand, Khozema, on the Flex contact center business? Because I was thinking back on this one, it had been landing some sizable wins, I think maybe 9 to 12 months ago. I know you can't mention every product in every single earnings call, and maybe it's just a function of that. But there -- there was mention of a company had rolled out Flex to over 15,000 agents, right? So that's a very, very large contact center. Is -- I think, in my mind, it may be felt a little quieter lately, maybe you're just been emphasizing some of the newer and more AI-oriented products. But any perspective on the demand and adoption out there for Flex?
I think it's largely a product of what you just said, Mark, like we don't try to turn every earnings call necessarily into like the greatest hits. But I think the -- there's 2 things, though, 2. One is we also took the Flex product and put it inside of communications, right? So previously, it had been part of this data and applications business. We thought that moving it back into communications was a better move. The reason being that what we were finding was inside of the company, there was channel conflict, that's usually a bad setup for a customer. And so now a customer can buy kind of the entirety of the voice stack, whether that's just purely IVR workloads, it could be our task router capability, which is a separate SKU, it could be our studio capabilities, it could be a full-fledged contact center. So that continuum is now what we offer. So that's kind of one part of it. I think the other part of it is that -- it's kind of the conversation that we had a moment ago with respect to Copilot. So we have introduced this Copilot. I think what that's going to accelerate over time because I do think it's going to take some time is that there's like a real transition that's happening inside contact centers right now, right? Where AI can do some really, really interesting things, certainly, initially in terms of better informing the context of a call, better deflection. But I think over time, ultimately, what probably starts happening is, is that there's a reduction in agents and those values start to accrue to AI and stuff like that. We're not close to that yet, but I do think that will happen.
Well, I may come back to that in a moment, but we have about 2 minutes left. I want to just check and see if we have questions from the audience. If you do, go ahead and raise hand, and we will run a microphone out to you. Okay then. Khozema, let's ramp on exactly that. Let's ramp on the topic of AI and seat compression. If I put it to you this way, Think about the imprint of Gen AI technologies on the contact center landscape, project this forward the next several years because as you've said, these technologies are making a big impact, right? As you outlined with the University in Mexico. Do you subscribe to a view that the LLM has become more powerful? They're going to deflect more calls? They'll answer more questions? And that eventually, do you start automating away some of the human personnel in the contact center?
I think eventually, that's inevitable Yes. I don't think it's tomorrow. And I think, look, the LLMs are already becoming like way more powerful. I'll just maybe edit one thing, which is they require context, right? So a contact center experience is pretty lousy if all that's actually happening is that they're sending you and I to self-serve right? Like we're still frustrated at that point, right? And so nothing's really changed. We've just gone to a different place. I think what's important is you need rich context, which is really where our segment and our unified profile comes into play. I think that you want the AI ultimately to solve the customer's problem. And so I think even if there's a reduction in seat count, the value that you were previously generating, let's say, from a license-based seat arrangement probably transitions now to a solved problem AI-oriented workload. But I think the other reality is, look, I mean, a lot of the industries that we serve, like let's just take financial services or health care, right? I mean you operate in a financial services industry like it's very regulated, right? And so it's not like they're going to flip over to AI immediately. And I just think that, that process is going to take some time. It's different in, perhaps, the university environment. It's definitely different if you're ordering a pizza. I think if you need a financial service or health care outcome. Man, it better work, it cannot hallucinate, and you've got to ensure that the safeguards are in place.
We can't be getting incorrect medical advice from an LLM.
Exactly.
All right. On that note, Khozema, I can't thank you enough and wonderful to have you here.
Thanks.
Read the full transcript via the API
You're viewing the first half of this call. Get the complete Twilio Inc. transcript - plus 252,000+ transcripts from 12,000+ companies, speaker segments and full-text search - through the EarningsAPI REST API or hosted MCP server.
Get an API key View API docs →For developers and AI pipelines
Programmatic access to Twilio Inc. earnings transcripts and 252,000+ others is available through the
EarningsAPI REST API and the hosted MCP server.
Quarterly plans from $105 - full transcripts, speaker segments, full-text search,
and the /api/v1/transcripts/recent polling endpoint for ETL pipelines.