We have a packed agenda for you today. We're a little bit behind schedule. So here you can see we're going to go, Kelly's going to take us through finance and strategy first. Then we're going to have our product leaders come in and talk about AI. And then Chris Morrissey, our Head of Contact Center, is going to talk about the Contact Center progress we've been making. And then we're going to bring back Kelly and have a panel with our executives. So when it comes to Q&A, please save all the finance-related questions for the executive session at the end, since Kelly will be back on stage. And we might have some time for a few questions in the AI and the CX sections.
For everybody, just please, if you want to ask a question, raise your hand and wait for me to bring you the microphone, just so that the people in the virtual audience can follow along. I think we can go to the next slide. All right, a few housekeeping items. We're going to have some non-GAAP financial measures. See our appendix for reconciliation to the GAAP measures. Next slide. We'll make forward-looking statements. Please be familiar with the Safe Harbor Statement and also see our risk factors in our 10-Ks and 10-Qs filed with the SEC. Without further ado, here's Kelly to tell us a little bit about finance and strategy. Kelly.
Okay, thank you, Charles. Before we dive in, though, I want you all to meet Michelle, Zoom's new CFO.
Hi. I'll be brief so you guys can get to the content, but I thought I'd just do a quick introduction, and then I think I'm going to...
Yes, do you mind coming here, though?
No, no problem.
Everybody who's remote can see you as well.
And then I'll come back and maybe talk briefly, but go easy. It's day three on the job. So just a little bit of background about me. I started out as a CPA one day, way back as a consultant at Arthur Andersen, but really I spent the last 25 years at Microsoft on its commercial business, from roles that went from M&A to partnering with engineering to sales and marketing. Kind of the red thread of, I guess, my time there was growth through transformation. So I am super excited to be joining Zoom in particular at what is, I think, a very key kind of pivotal moment in their history. And so we'll be talking more about that, I'm sure, with Kelly, and then I'll join you back later. But I'm excited to meet with all of you again at earnings. So see you then.
All right. Thank you, Michelle. And yes, you all have plenty of time to spend quality time with Michelle in the near future. So hopefully all of you heard Eric's keynote this morning. But what we really want to highlight is all the amazing new products that were announced today. Some of you, I already heard your excitement about custom AI add-ons that are being paid. So that's super exciting. But even before that, I think what's really great to see is AI Companion 2.0, which is really your personal assistant that has context and can start to take tasks from you. And that is what's going to continue to be included for all users at no additional cost. And then we are starting to monetize some key AI functionality, including Custom AI Companion, as well as vertical solutions for healthcare and education.
And then last but not least, one of the key announcements this morning is how we are continuing our foray into the frontline worker arena. After the success we have seen from Zoom Phone, as well as Workvivo, especially in the momentum we've seen from them over the last couple of quarters, we're really excited to address this 2.7 billion-person market that has largely been underserved. I heard a stat recently that it's 80% of the workforce and 1% of the SaaS spend. So I think that really highlights there's a significant opportunity here. Thank you. So as the platform is expanding, so is our total addressable market. So you can see here that we expect by 2028 to have an addressable market of $235 billion. And this is through a very deliberate strategy around organic and inorganic expansion.
And includes all. You can see that $56 billion up there at the very top is from all those new products that were announced this morning: Custom AI Companion, frontline worker, healthcare, and, sorry, AI for healthcare and education. And then, of course, Workvivo, $32 billion. So that's also really exciting to see. And then, again, we talked a little bit about the frontline worker, but I think that is an area that will continue to expand and stretch our TAM over time. We have also made considerable progress with our customers' expansion on our platform. So what this shows you is that customers with three or more products have grown at a CAGR of 15% over the last three years, from FY23- FY25, and now contribute 66% of our annual recurring revenue, which is just amazing.
I think what's even more amazing about that is you can also see that it has reduced churn for these customers by over 50%, so what that's creating is this very resilient and growing stream of revenue, and of course, the lifetime value of these customers just continues to grow. At the bottom of this slide, what I want to highlight, though, is there is also considerable opportunity remaining in our existing install base. This is a stat we update for you annually, which highlights that at the enterprise account penetration level, Zoom Phone and Zoom Rooms penetration is, they're each still under 20%. So tremendous opportunity exists there. All right, so let's take a look at some examples of how this expansion across the platform actually occurs with our customers.
The first case study here is a large Southeast bank that started in FY21 with hybrid work for meetings and rooms. You can see they loved Zoom so much that in the next year they added Phone. They continued to go on and do a pilot of Workvivo, and then by FY24 had expanded both Zoom Phone and Workvivo. And over that period of time, we saw a three-times, almost three-times increase in their ARR. The next one is a North American logistics company. So this company is also really exciting. It came to us again in FY21, as many people were looking for amazing solutions. So they started with meetings, you can see here. As they've gone through the years, they've added Phone. They've also expanded meetings. They migrated to the Zoom Workplace bundle and then continued to add Workvivo. And look at this one.
This one has grown 22 x since its initial MRR ARR purchase. And I think that just highlights for you how the strategic land expand motion continues to work across Zoom. And especially as this platform is expanding, the team has really embraced this, and our customers have, and their commitment to the platform. And then I love, love this story. So this is a very unique case, as you all probably know. Typically, when a company is acquired, the acquirer gets to dictate what services and vendors are going to be used. This is a very large tech company that you all know very well that was a large Zoom customer. And when they were acquired, they were very upset about the prospect of switching from Zoom to Webex. And luckily, their acquirer saw the benefit in keeping them happy and listened to them.
And not only did they migrate and let them keep Zoom, they also migrated the rest of the company onto Zoom and expanded beyond that to Zoom One Enterprise, one of our bundles, Contact Center, Events, and Webinar. And on a combined basis, that spend has grown 50% since the acquisition. As we continue to see the strategy of land and expand working for us, we see this growth being reflected in our larger accounts. So on the left, you see your customers with greater than 100K of ARR, and it's grown 25% since FY21. Even more impressively, customers with greater than a million of ARR have grown 33% over that same period of time. So again, this just highlights the resilient revenue, the increase of lifetime value that our ever-growing platform is, how our customers are contributing with the ever-growing platform.
Customers are also showing their commitment to Zoom by lengthening their billing cycles and their contract terms. On the left here, you can see we've had a 27-point improvement in the percentage of revenue that is on annual billing terms. That's what you see in the left. So sub-annual in Q2 was 76%. I'm sorry, sub-annual was 24%, annual was 76%, as compared to 49, almost split 50-50 four years ago. And our average contract length has improved by seven months, sitting on an average of 18.1 months. This is all customers, so not just enterprise. We often talk about enterprise being around three years. This is enterprise and online, so direct and online customers combined. We continue to have a very diverse customer set, both geographically and across industries. So you can see that Japan has now become our top international market.
Even with the currency headwinds we've experienced, we continue to see tremendous opportunity and have had great success there. When you look on the industry side, we've had consistency in the top couple of verticals with financial services and technology. But given the investment that we've continued to make and even are continuing to make with AI now around healthcare and education, we've seen those two move up. Thank you. This is a slide for those of you who are modeling around our renewals. Just want to highlight that while, as a reminder, Q1 is still heavily weighted towards renewals, but you can see from 2023- 2025, we're starting to see this slowly but surely balance out.
I mean, it's going to take some time that we ever get back to a period where Q4 is a large renewal that you typically see in SaaS, but you can see it's starting to even out over time. We have continued to have an amazing track record of profitable growth. I mean, looking at this slide, it's really incredible to remember that in FY20, our revenue was only $600 million. So we've come a long way in a very short period of time. But over that time, while our focus is absolutely on re-accelerating growth, we remain committed to providing very high operating margins and free cash flow, and I'm sure you're all very familiar with the FY25 guide of 39% from a non-GAAP operating margin and 35% from a free cash flow margin.
AI is obviously the number one priority in the company right now in terms of investment. You heard a lot about that this morning. We've heard from a lot of you concerns about how we're prioritizing this without having significant impact on our margins. This is how. First of all, we take a federated approach, as a reminder, which means that we are not tethered to one single model. We leverage models from OpenAI, also Anthropic, and our own model. That really gives us the flexibility to have more accuracy in a quicker time at a fraction of the cost compared to our competitors. The team is also and has been the whole time I've ever been at Zoom focused on ensuring that our DevOps and operations run as efficiently as possible.
You can see on this left-hand graph what this is highlighting is how even in the last two years, traffic has shifted from the public cloud to our colos. So currently, we're running at 91% of our traffic in our colos. As a reminder, that's a much more efficient way for us to process traffic. We have 32 data centers today on a global basis as compared to 27 is what we had two years ago. And that's helping provide the opportunity to move more and more traffic into those data centers. The way that we do that initially for any products is they get launched into the public cloud. That's why you're always going to see some traffic and some commitment in the public cloud. We see and assess the usage to make sure that we have enough capacity for them.
And then once we understand the usage patterns, as quickly as possible, we bring those into the colos. What that is doing for us is driving economies of scale. As you can see now on the top right graph, that just highlights how AI Companion COGS per active user has come down over the last even two quarters. You can see we've seen a decrease by over 30%. So this is great. This is a number one focus of the team so that we can continue to democratize the use of AI across all of our customers. You heard it from me this morning saying, "We want everybody to have access to this." And yet our commitment to all of you is we're doing it without significant impact to our gross margins, which you can see at the bottom.
The FY25 guide is 79%, down only 100 basis points from where we were last year. Okay. We've also continued to drive commitment to free cash flow. The improvement you see here from FY23- FY24 was really all about improving operational efficiency in the company and in our DSOs, and then the growth from FY24- FY25 was really around expansion, expansion of billings and expansion of interest income, and I think that's really great to see as you don't always want this to just be from driving efficiencies. You want to see it from expansion and improvements in other areas, so we're really excited about our 35% guidance for this year. In terms of capital allocation, I know this is a question that many of you often have, and we just wanted to remind you that we're taking a balanced approach to capital allocation.
We are focused on maintaining flexibility in our balance sheet to pursue potential organic or inorganic opportunities as we see them, but also returning cash to shareholders. So what you see on this slide is the various ways that we've used capital over the last several years. As a reminder, in FY23, we completed a $1 billion buyback, which more than offset dilution in that year. We are currently under a $1.5 billion buyback authorization. And what you see here is what we've executed through the first half of the year, $438 million, which almost offset the dilution for that period of time.
And we're continuing to execute in Q3 and Q4. Okay. Something you've all been waiting for. I know you're dying to see the new long-term model, right? So this is our long-term model that we're sharing with you. It's been a couple of years since we've updated this.
We just stepped through each of these components really quickly. Our gross profit, we are maintaining 80%. Again, slightly under that for this year, but we certainly think for the long term, with all the efficiencies being driven, that we will stabilize that back again at 80%. With the prioritization around investing in AI and expansion of the platform, we are taking up our long-term model around research and development. Previously, we added at 10%-12%. We think that a better level of spending and investment there is 12%-13%. We have really focused on driving efficiencies across other areas of the model, though, including sales and marketing, where you can see is actually coming down from the previous model at 30% to 25%-26%.
And this includes some of the efficiencies we drove through the team last year and very focused go-to-market strategies on a global basis. And then we are always very disciplined around our G&A spend and are very proud to take it down from 8%-10% to 7%-8%. This results in an operating profit range of 33%-36%. This is a model, a framework that we will continue to use as we evaluate opportunities for organic and inorganic growth. As we've always said, the number one priority is re-accelerating revenue. And to the extent we see the opportunity to do that, we would certainly take down margins, but this would be the guide as to how we would accomplish that. The other area we've been very actively focusing on collectively as an executive team is managing down stock-based comp and dilution.
The first graph at the top is stock-based compensation. Some of you are very familiar with the fact that we eliminated our top-up grants last year. These are a lagging indicator of expense. Even as we make changes, it takes a while for them to move through the P&L. As a quick reminder, a quarter of our stock-based compensation expense is attributable to top-up grants in FY25. That's down from a third of it in FY24, and that will feather off all the way through FY27. What you see here is kind of the range. FY24, we're at 23%. We expect for FY25-FY 27 to be in the range of 21% improving to 15% over that window of time. For FY28 and beyond, we expect stock-based compensation to be under 12%.
And then from a gross dilution perspective, excluding any buybacks, we have also, as a company, taken actions and implemented compensation structure changes that were announced here in Q3. And these include things like bringing our new hire grants to market. We are also implementing a tiered bonus structure. Some of these are happening over time, so it's going to take a couple of years to have this all fully move through. The other thing that you see represented when you look on the timeline at the bottom is that there are a significant number of employees that were hired in FY 2021 and 2022 who are hitting their four-year anniversary in FY 2025 and FY 2026. And as a reminder, the way that we implement grants here is you get a four-year grant at the beginning of your tenure. We don't give annual refresh grants over time.
We give another four-year refresh grant at the end of that period of time. So there's going to be some impact for FY25 and through 2027. That's why you see that increase in the range of 3.5%-4%, while we see those employees come up on their four-year anniversaries. But by FY28 and beyond, we expect our gross dilution rate to drop to below 2.75%. Okay. I feel like I talked super quickly so we could get us back on track. But what I just want you all to take away from, most importantly, the keynote today is that we're very proud of our established leadership in our core video, but very focused on our AI innovation as well as the tremendous momentum we've already seen in Phone, in Contact Center, and in Workvivo.
And while supporting all of that, the commitment from your finance team here is that we will continue to focus on driving profitable growth and maintaining our strong balance sheet. Okay. I think I got us back on track. Right? Two minutes. Okay. Great. So there we go. Now I would like to introduce Jeff and Bo, and they're going to come up and give you some more information about our AI vision. Thanks, Kelly.
So I'm Jeff. I lead product for Workplace AI Meetings and Spaces. And my partner in crime here, Bo.
Buddy, buddy.
Yes. That's right. Head of Workplace AI or AI overall. Can you go to the next slide? Yes. So you heard this in the keynote today. We have built an AI-first work platform. We're really focused on that human connection. Why do you have AI? So that we can focus more on connecting as people. I'm going to reframe this a little bit because we've got an investor audience. But thinking about first, so three things. First, we're producing end-user value, things that will help us in our daily work. Second, how that value is packaged up in a way that delivers customer value. And then third, how that is developed efficiently so that we can return some of that value to investors. All right. You can go to the next one. So what is our approach to AI? Kelly mentioned our federated approach.
We're taking the best of what's out there, and we're packaging it up in a way that users don't have to deal with that complication. So simple, easy to use user interface, leveraging the best of what's out there in AI. It's also platform-wide. So it's available wherever you are in the Workplace application. What that means is I have a personal assistant that has one brain. It remembers what I talked about when I was on this other part of the app. I don't have to find an assistant for that particular task. I have my personal assistant that's with me all the time. And we are always focused on doing this in a responsible way that takes as much care of customers' data as they take of their own. So go to the next one.
We've seen really strong adoption of our AI tools to date, 40% increase in active users, really, really strong. That's just in the last quarter, and you saw this data as well in the keynote, 57% of the Fortune 500 have enabled AI Companion. And we're doing it at high quality, so any of these applications, it doesn't matter if you have the feature unless it produces a good result, so really looking at the core baseline capabilities of transcription and summary as a measure of how all of the features that we build on top of that function, so 36% fewer errors in transcription, 15% in summaries. Go to the next one, so what is AI Companion 2.0? AI Companion 1.0, by comparison, was a set of specific features. AI Companion 2.0 is that single-brain personal assistant that goes with me wherever I am in the app. It's contextual.
So if I'm looking at something in the app, my assistant knows what I'm looking at and can answer questions relative to that. It synthesizes information. So whether that's specifically, "I got a big long document, just give me the TL;DR," or "I've got a million messages, tell me what's my priorities," those are the types of synthesis tasks that AI Companion can do. But it doesn't just summarize all that stuff for me and make me do all the work then. It can actually take action. So this is where we're going to be developing a ton of new features where AI Companion can do the work for me. Already really capable at generating documents based on a random conversation or helping me compose something in a voice that's more professional. So we'll be building a lot on the takes action side.
This is what it looks like in the Workplace app. So you'll see AI Companion as a new side panel. Again, really easy-to-use interface with the conversational UI that's always there. And you heard XD mention how we're pairing the conversational UI with the graphical UI. So the things that make sense when you want to be presented with something visually, we'll do that. The things that make sense when I want to have a conversation about it, we'll do that. And they work together. Really focused on the core of where our users are today. One of the challenges that people have with AI is, "Okay, a lot of stuff that I can do. What do I do first?" So for users of meetings, which is almost all of our users, we start with meeting summaries.
How do I summarize all of these discussions that I'm having in a way that's now digestible? How do I have better meetings? I can create an agenda for some of those meetings, bring some structure into those meetings. How do I follow up effectively after those meetings are done? Building tasks, the task service that we've built is really focused on those AI-generated tasks that are coming out of meeting summaries. How do I follow up effectively on those? Not building a really complete project management solution. The things that we build are really focused on advancing our core in the places where our customers are today. What do those tasks look like? Really simple, really easy to use.
When my assistant says, "Hey, you've got these three tasks that came out of that discussion that you had in the meeting," now those become objects with attributes. They can be assigned to people. They can have a due date. They can have reminders. So instead of, "I have a meeting. We kind of forgot about what we talked about, and we have another meeting to figure out what it was we were supposed to do in the first meeting." No, we have a meeting. We do the stuff that we talked about, and then we have another meeting that's really progressing the project to completion. All right. And it's not just about virtual meetings as well. So some of my other products are all about the spaces or things that we do in the office. In-person experiences are just as important, if not more so.
So AI Companion for in-person meetings, meaning that I can have my assistant take notes, summarize, create action items for a conversation that we have face-to-face. As well, making it worthwhile for me to come into the office is all about who's going to be there, where are they going to be, are we going to be working on something that's valuable, or am I going to be around a bunch of people that I don't work with every day? So our AI Companion can make sense of all that as well. The more flexible that our workforce is, the more challenging it is for them to come together and make valuable time.
AI Companion for Workspace Reservation will help us say, "Okay, I should be in the office on Tuesday because there's a really core group of people that I work with, and they're sitting on the fifth floor." All right. Bo, I think I'm going to turn this over to you.
Okay. Thanks, everyone. XD was talking about technology. I just want to recap that quickly, see how we can achieve. You saw that data, right? 36% accuracy or error rate reduction and 15% error reduction for the action items. How do we keep up with competitors going forward, right? Our winning recipe is really federated approach. We're talking about we federate the best of the best. We have Meta, we have Anthropic, we have OpenAI. And we build our own small language model to optimize not only for quality but also for efficiency and cost for that purpose, right? We know that AI Companion is offering at no additional cost. How do we keep that? You can't just blindly use one model, but rather than use the right model for the right task accurately and also cost-efficient. So this is the key thing that we're looking into.
And the next one is really, in addition to the general model, we also provide customization capabilities, which is introduced this morning as custom add-on. Go to the next page. Yeah. If you think about custom AI, we're not really talking about building everything from ground. When you customize it, what is a customize, right? Fine-tuning. We can use prompts to do the fine-tuning. We can ground the conversation with your data. That is part of the fine-tuning, customization. You can have special dictionaries that you can recognize, special terminologies for your organization. All those things are really customized. Not only that, but you're thinking about from a data perspective, not only we're handling Zoom data, we're also handling your company data, right? And retrieve that information. You can bring in, bring your own data into the AI Companion.
And also, we extend it beyond those two buckets into the third bucket, which is some of your data stays in your line of business application, like Jira, Asana. And through the application, we can extend search knowledge beyond just our Zoom data. And furthermore, I think we talked a little bit about federated through Perplexity. I had a meeting with an executive yesterday. I asked the question, "I'm going to have a meeting with this person. She's the VP of blah, blah, blah, and help me to prepare that." And guess what? I never met this person before. It's search internet and find that information. Give me suggestions. What he is interested about. Isn't that great? This is the power of AI Companion. Go beyond just the Zoom data, right? And then the personalized, we're going to talk about that in the next slides. Let's go to the customized.
Yeah, this is an example. I think Smita talked about this workflow. And I want to call out a couple of things. Number one, with the third-party application extensibility, you can interact with your AI Companion reactively, ask information. It's going to pull that information, as I called out earlier, from Workday, right? And you can attach that, continue the work, the action, right, into ServiceNow. And then the third one, if you notice that the way the prompt shows you that AI Companion understands the context, proactively trigger that. Do you want me to do that? That is what we call action-oriented information flow, right? Not just to give you a bunch of reminders. Go beyond that and do things on your behalf, thinking on your behalf. And then you can say, "Yes, guess what?
The whole workflow, the task is completed." And combined with what Jeff just mentioned, task services, it's critical. That's a single source of truth. It goes across all the other applications like Jira or other applications. All you need to worry about is, "Are all my tasks completed, on track, or anything unfinished?" As an engineer, we all know that. Every day, we just have a list of things we need to check. At the end of the day, we want to make sure how many we cleared, checked, and move forward. That is why the AI Companion, I think, 2.0 is the critical part we're thinking to that direction is also AI-driven information flow, action-oriented information flow, really go beyond the reminder to do actions. And we're going to do more such kind of actions moving forward. Next, please. AI Studio. People are asking, "What is this?
What's the relationship between Custom AI Companion add-on, right?" You get add-on license. You're able to use AI Studio. AI Studio is the tool to allow you to further customize and build the skills for your company. And this is the one that we feel like really unleashed the power of the AI. You can use this one to bring your own data, customize it. You can use this one to customize meeting summary that we talk about. The general meeting summary today is very useful if you missed the meeting. If you are in a meeting, you probably have 80% chance, I have to be honest, right? You're not going back to look at the meeting summary. But guess what? It doesn't matter because you can customize that in a way.
We're going to build this capability to automatically remind you, "Hey, this sounds like a one-on-one at some point," or interview. Do you want to generate a summary based on that template? They say, "Yes." Awesome. No, this is not an interview. This is a career discussion between me and my directs. Use the other one. Guess what? The report, the summary will be tailored to the things you specify. And that is a customization capability, right? Next, please. I can talk. You can talk.
Oh, sure. Yeah, so looking at some of the personalization capabilities, first, personal coach, I'm really excited about this one. So some of the things that we'll be able to do is look across all of your communications. And you can set goals for how you want to improve those communications. For me, I want to have more inclusive, professional meetings. I want to have meetings that are more structured and have an agenda where we accomplish things. Those are things that I can set goals on. AI Companion can say, "Here, how's your tracking relative to that?" And I can improve over time. Very, very excited about my personal coach. And Custom Avatars help me scale my communication.
So today, I have to either be there in person to communicate something, or I have to communicate with dry text, or today, I can record a live video clip of myself. With our Custom Avatars, I'll be able to have a lifelike version of myself based just on a text script deliver a video clip in any language. So I can come in. I can speak Mandarin. I can speak Japanese, whatever. I can speak to a global audience in their native language with my voice and likeness based just on a text script. It's really, really, really powerful. And the first time you see yourself speaking a different language, it's really strange. But yes, super excited about that one as well. As well, building vertical applications on top of this customization theme. So remember, this is number two in the three things we're supposed to remember today.
First, that we had user value. I think we've been able to demonstrate that. Second, that we're packaging that up in a way that we can monetize that and produce both customer value and value we capture. And then Bo talked about the federated approach and the ways that we're managing costs so that we deliver that efficiently and we can return shareholder value. So I think.
I think I want to add one comment about EDU sector. I don't know if you caught the demo Randy presented in the keynotes. It is really not about not taking notes, but really modernizing the teaching and learning. Just think about this one, right? As a student, you're taking the lecture. You're taking notes. But with a new Workplace for EDU, you can mark the transcription. You can highlight. You can ask questions with AI Companion. At the end of this one, end of the session, AI Companion is going to generate tailored pop quiz for you or notes for you and help you to do more research. Same thing for the educator. The professor knows what questions students are asking in the classroom can optimize for their future preparation. This is really about revolutionizing EDU. I'm so passionate about it. I think this is a great thing.
It would benefit the higher education industry, benefit the professors, and also the students. I can't help but jump into talking about it.
Yeah, no, I'm building on that. I think both for knowledge workers as well as for the EDU space, it will not be very long before it seems ridiculous that we are manually taking notes. It will be a thing of the past, just like writing with pencils on paper. It is a tool that will be a baseline tool for everyone to use. And it is coming faster than I had ever anticipated. So yeah, so my seven-year-old can thumb-type better than anybody I've seen. So yes, there is obsolescence in the future from this technology. For clinicians also as well, being able to focus on that human interaction, having an unstructured conversation about their condition, that is one of the things that AI is just fantastic at.
Taking these unstructured conversations that we have, might be rambling, might go over here, might go over there, and turning it into a very, very structured summary of what happened, what was important, and what you need to take away with the next step. I think that's going to be revolutionary for healthcare as well. So just a reminder of the breadth of the Zoom solution. We talked a lot about AI. In this diagram, we show that at the top because it's infusing every aspect of the platform. From the Workplace side that I tend to focus on, that's focused on our knowledge workers, but also on the business services side as well, underpinned by an open ecosystem of partners. So hopefully, you kind of get the idea of the diagram there. I think that's it. Yes.
So before Chris goes on, I think we have about 10 minutes for some Q&A for Bo and Jeff. If anyone has any questions, just raise your hand and I'll bring you the mic. Meta?
Maybe on the education piece, I would assume that for the most part, there's a split of in-person students and those who are remote. And so how do you assure that kind of the ones that are in-person are getting the notes? Are they in a team flow where they can get kind of that same amount of information? And then maybe just on the healthcare piece, I might have misunderstood on the clinician piece. Are they recording a patient conversation or? Okay.
Yes.
Is a patient opting into that?
Yeah, absolutely. Absolutely. And we do take a lot of care in these in-person experiences that we're talking about with taking concern over user privacy as well. So first, looking at that hybrid environment and how does everybody get an equal amount of kind of presence in the total experience. So we are encouraging people, even if I'm a student in the classroom, to connect to Zoom. And that is not for the audio-video that is the core of why people would think of connecting to Zoom. It's for all of these ancillary features. So do I see the live summary? Can I interact with the asynchronous chat conversation? Can I ask questions? Can I raise my hand? All of that puts everybody at an equal playing field. So connecting with their personal devices in the room as well.
Yeah. And also, we think that for the EDU part, it's going to be hybrid, continue to be hybrid. And once you have hybrid, you have to turn on the meeting, live meeting or the Zoom meeting. And the students can join that, even if it's in the classroom, and take the notes, highlights things, and that can interact. So that's more like a consistent experience, really, across in-person and also hybrid.
We're seeing that in EDU as well, where even if it isn't part of the professor's plan, that people are, I'm a student. I bring my device. I'm turning on a recorder. I'm having an AI note taker in the classroom anyway. But having that be a part of the interactive curriculum really brings that into where everybody can be included rather than we have people that have one set of tools and others.
Here's a mic. I can talk more. He just handed it over.
Okay. Got it.
Siti with the Mizuho. Quick one. On the AI Companion, I saw $12 per user per month and the new products today contributing $56 billion to the TAM. Is that $56 billion coming from the $12? How do you build the TAM from that? Is it enterprise, online, consumers?
That's not one we're prepared for.
Yeah, I think that's more of Kelly at the end. Kelly will be back on stage for the executive Q&A, and we'll talk to the TAM. Darrin, do you have a question?
Appreciate it.
Hey, guys. Thank you for the presentation today. Look, I wanted to ask kind of a big picture question that really applies to many software companies today, which is that any medium or large organization, I think we can safely say today, is going to have a lot of different software platforms that they're working with, right? And most of those guys are working on coming up with some version of Copilots, agents, whichever kind of term you want to use most recently, right? And I'd love to hear just kind of how you imagine Zoom's AI capabilities, I guess, fitting into that kind of larger ecosystem. You talked about the Custom AI Companions being able to kind of link out to other services.
But just maybe paint a picture for us of are you hoping that kind of Zoom's tool would be more of an orchestrator across some of those other tools? Would it play some different role that I'm not imagining in terms of how it integrates? I'd love to just hear some more about that. Thank you.
I'll take one shot, and then I know you've got some points on this as well. So first of all, where is the core use case for Zoom's AI Companion? And that's in these collaboration-centric use cases. Centered around meetings, we know that people are having meetings in Zoom today. You have this anchor of that experience. Anything tangentially related to the content is where you're going to want to have AI Companion from Zoom as the core user interface. As we start to expand into the rest of the platform and people are more engrossed in the. I've got the diagram up here. In the rest of the platform, it starts to extend into other use cases. So we can see an orchestration from AI Companion as the user interface that people are accustomed to into a workflow.
So connecting from meetings into a collaboration workflow, extending into the ecosystem where needed. So definitely see that as really the core use case for an AI Companion from Zoom.
Yeah. On top of that, actually, you think about how it can access information across the Zoom ecosystem, right? How do you, while you're doing things with your Zoom AI Companion, and you have Outlook calendar, you have a Google calendar, you have Outlook email, Zoom Mail. The one that I think in the basic AI Companion today, we support M365 , or we call that we support Microsoft and Google. So you bring those information in. That's kind of construct your main knowledge when you try to prepare a meeting, when you try to drive meeting summary. The other part is interesting is thinking about all the meetings, all the information is you can export actually also the meeting summary or meeting transcript into VTT format that you can input that into Zoom, and Zoom can take that intelligence and continue the rest of the things.
If you want, if you trust, you believe that Zoom AI Companion is your personal assistant, truly personal. I think one of the things, I don't know, when we talk about Zoom 2.0, Jeff mentioned that is, look, Zoom's vision is one AI Companion, one history, one memory, right? It has a full context among all the Zoom apps, all the things here, information is floating inside, and Zoom AI Companion has access. Does the Microsoft Copilot have access? No, they're very siloed. I just want to be I don't know. It's out of script, but I just want to share with my perspective on that. Because you think about Copilot for Teams, they never talked to Copilot for Word or PowerPoint. How can you get intelligence, right? We are the same, same AI Companion.
That's why we're able to orchestrate actions across and derive information from all the sources, not only within our ecosystem, but also beyond our ecosystem.
Yeah. As a user, I don't want to have to go find which agent. I need to ask which question. I want to have my assistant with me, and my assistant figures out where it needs to go for the information. So it's seen an army of agents running around, and it seems very confusing for a user.
Hey, guys. Alex Zukin from Wolfe Research. I was actually going to ask Darrin's question, but Darrin beat me to it.
That's a good question.
It was a great question. I'm trying to understand if you can encapsulate the main change in the innovation front, what the difference is between Zoom 2.0 Copilot or Companion, sorry, versus Zoom 1.0, and I ask that because it seems like there was a lot of product launching today. There were vertical AI Companions and education and healthcare. There was Tasks. There are monetization features and orchestration parts of Companions, so are these all separate bespoke products that require separate bespoke engineering and sales and delivery methods? Did you unlock, unleash something in the underlying platform that wasn't there with 1.0 that enables these things? Just help us kind of conceptualize that.
That's why Jeff and I work so closely. You can see where buddy- buddy is. I'm the engine. He's the car under the hood. I think that's a great question. One of the things we believe, comparing to 1.0 or comparing to Microsoft Copilot, by the way, I was with Microsoft for 16 years before I joined Zoom five months ago. It's a great company, great product. I think one of the things, a differentiator is number one is federated AI that we're building on top of the best of the best. So quality and also cost efficiency. The second one is Zoom AI Companion continues to be no additional charge for paid customers. That's really the second one. I think the third one is what we're talking about, the full context. Zoom AI Companion has one AI Companion, one history, and one memory.
And that thing is going to come on more and more into an impactful experience. Thinking about this, the general summary is not fitting your needs. How can you customize? How can you do that? That's the direction we're going, be more contextual. And comparing to 1.0, 2.0, and you can comment on that, I feel like one fundamental change in 1.0- 2.0 is really we're focusing on action-oriented information flow. It's beyond reminder. And 1.0 can generate summary, and in the 2.0 and moving forward direction, we're going to turn those reminders into actions and change our internal Zoom application services and also third party that you can empower to do more. That is the things that we feel like it's a huge differentiator comparing to 1.0 and 2.0.
And I'll add a little bit from the user experience perspective. So in 1.0, I could generate a summary for a meeting, and I could ask questions in that meeting about that meeting. Very limited context, and it didn't go anywhere else. So in AI Companion 2.0, my assistant understands a meeting I had last week, and I can ask a question today about a meeting I had last week. And all of that information in terms of memory and context is now broadly across the platform instead of siloed in a particular application or feature.
So it sounds agentic, and it sounds like it has memory. How much actual data infrastructure work is required by the customer around their data estate, around the integrations, around all these different layers to make this work?
From a data perspective, I think we do respect customer's decision to what to share with Zoom AI Companion, so for Zoom data, the infrastructure backend part, 1.0 and 2.0 is very close. We're not really completely redesigning the backend data infrastructure, but the thing is that opens the surfacing for AI Companion to access those information. As you can see that we're building the applications really natively integrate. One word you probably heard, AI first, right? What does it mean, AI first, and why we're not AI after or plus AI, right? Because we designed this one from the bottom, being able to chain them together and access those informations, so for information outside of our Zoom ecosystem, customer can choose to share calendar with AI Companion. If you choose that, we're going to do it for you. If you don't, sorry, I can't help.
So it's really up to user or customer to decide what to be included. Same thing as the third-party ecosystem. Thinking about it, you want enable information exchange or task completion with Jira. First thing you need to do is probably you need IT admin enable that. Allow you to use it because someone prepay license for that. And then customer needs to authenticate where you're using that first time you're using Jira. It's not like blankly all of a sudden information will be flow into AI Companion. First time you use it, you need authenticate. Once you authenticate, your AI Companion has the same access right as you. Then everything's streamlined and goes. I hope that answers the question.
Yeah. So from the AI Companion 2.0 perspective, it doesn't take a lot of infrastructure or configuration. It's just an enablement flow. So I want to access to this, this, this, and this, not this, not this. When we get into the customization, that does require admin or integrator effort to say, well, I want to build a custom skill, or I want to have this specific data source that I'm going to give AI Companion access to. So from that higher level stuff, the customization side, there is some effort on the base included offering. It's very little.
And also one more thing, I think last thing. It's interesting to see Zoom has open ecosystem. It's very open. We collaborate, right? While you're bringing your own data, people asking, I have a terabytes of data. What are you going to do? You're going to create things there? No. We support bring your own data. We also support you can bring your own index. A lot of enterprise customers have enterprise search. They have the index. We have partners that we can work with that. So in addition to bring your own data in a Custom AI Com panion, the AI Studio, you can also bring your index. This is on a roadmap, right? I just want to share with you to show that how open we are in terms of empowering, really address our enterprise customers' challenges, right?
Instead of reinvent the wheel and conquer the whole world with our stuff.
Thank you.
I think that's where the ecosystem comes in quite a bit.
Yeah. So they don't have to redo the work.
All right. Well, thanks both. Thanks, Bo and Jeff. And a lot of great questions. We also have Smita, our Chief Product Officer, which she'll be on our executive panel at the end. So if you have follow-ups on AI, be sure to ask them then. And without further ado, I'd like to welcome up Chris Morrissey to speak to us a little bit about the Contact Center.
All right. Oh, thank you. Thank you. Welcome, everyone. I hope you all had fun drinking from the fire hose today as we announced all of our products. I love it because I learned about everything we're doing, so just brief introductions before I get started. My name is Chris Morrissey. I'm the global head of our CX sales, go-to-market practice here at Zoom. A little bit of background about myself. For better or worse, I've been in the Contact Center space for the better part of 25 years. Starting back in the early 2000s, I was part of a startup, and when I say startup, I mean two of us. We built a Contact Center platform; it was a recording quality management program. Fast forward to 2016, we sold our company to NICE Systems.
Frankly, we were taking their market share, which is the best way to get bought by a small company, and then for the past eight years, I was at NICE running different sales go-to-market strategies for the US, and then the last two years at NICE, I was running it for the EMEA and APAC regions for the CXone product specifically, which is their cloud CCaaS product. Started at Zoom in November of last year. Have not looked back since, only because I'm way too busy running forward on everything we're doing and super excited about what we have going on right now, so I'm going to show you some of the go-to-market strategies, what's happening in the market, and why we're positioned to win, so when I first started, there were two key things that I wanted to focus on from a go-to-market perspective. First is our goal.
What's not exciting is to say Zoom is super excited to be the 13th CCaaS vendor in the market, right? I mean, that doesn't excite anybody. So we had to figure out a way to change that. And by the way, the market was already in place to change it. The second thing is we couldn't peanut butter spread across 74 initiatives and hope we actually made some meaningful movement on those. So we picked four focus areas that we thought would move the market and get better market share at a faster rate. So in no particular order, the first one was a focus on the international opportunity for a couple of reasons. First, international market is the market that typically moves to cloud slower than the U.S.
So when you talk about how many agents have moved from on-prem into the cloud, it's a little bit faster adoption in the U.S. The U.S. is by far the biggest market. But when you get in the international, a little bit slower. And then also making sure we had prime international target markets. We don't want to go after 500 countries and go try to be a CX expert. We know where the agents sit. We know where the biggest adoption is. We know where the biggest potential is. So we focus on some prime countries. And that means we have to invest in things like data governance, keeping data in region, language support, all those things we focus on. That all fits into the international focus. Second one is channel growth. And some people ask, like, why are we focusing on the channel?
Don't we have really good sellers that can do this ourselves? Absolutely. The challenge is when you look at a lot of existing customers on a Contact Center platform, they have 10, 20, 30-year relationships with partners already that were building out their ecosystem because they had Avaya and Cisco, and then they had some security, and then they had Salesforce, and they had all this premise stuff. And brands didn't want to deal with it themselves. So they got a trusted partner. Those are the partners we're signing up. So large VAR business partners, Cisco partners, and allow them to be a trusted advisor. So when the brand says, "I need to move to the cloud, where should I go?" They say Zoom, and that's where we get our business. The second thing is they are service. A lot of these companies are service-driven companies.
So they love products that they can wrap service offerings around. And CX is very, very complicated, and there's a lot of services you can wrap around that. So focus on the channel growth. Third area is the move-up market. And if you look at the trend again, when we say 40% have moved to cloud so far, premise agents to cloud agents, and analysts have different numbers, 40, 30, 50. I pick 40 because it's the middle of the road. So 40. Most of those are in the SMB market. Most agents that haven't moved are in the upmarket space. So if we're going to be there when that wave starts moving, we have to focus on upmarket and what that means from a product perspective. And then the last thing is taking advantage of what we're calling the AI-first total experience.
This is where, back to that statement, we don't want to be the 13th CCaaS vendor. There are dynamics happening in the market that make it very, very well for Zoom to be the leading vendor in this new dynamic as a disruptor, so you go to the next slide. This new trend, AI-first total experience, as the name implies, there's two aspects to it. The first one is the total experience. CIOs don't want to view customer experience, employee experience in silos anymore, both from a technology perspective, right? I don't want to have to go manage all these different products to really just provide communication services.
And the second thing is, if you wanted to provide a really, really good customer experience, and you went to a brand leader and said, "Who in your organization is accountable for delivering customer experience?" I have never heard a leader say, "It's just my agents," ever, right? It's usually everyone in the company or most people in the company. But the reason they haven't been able to do that is because technology has lived in silos. It wasn't a business strategy ask. It was a technology thing that put them there. So that's changing. Next slide. I'm sorry. I was speaking ahead of my slots, but this is the one who's accountable for total experience. This one. The other thing, as the name implies, what's happening in the AI-first total experience. AI is transforming the CX space, probably more than any space.
I know finance, healthcare, but Contact Center, CX is being innovated so fast. It's one of the fastest adopted technologies in our space. It's changing everything, so this is around predicting what's happening, automating tasks, and making it like a human-like or more natural interaction. This is where people, even if they're on-prem, the conversation is always starting around AI. Yes, they want to move to the cloud, but help me understand how is AI going to help transform my business? Because no one really wants to say, "I just took my premises stack and I moved it in the cloud and I didn't change anything," but I moved to the cloud and I totally transformed how my customers and employees work together to provide a differentiated brand experience, so this is a wheel you've probably never seen before.
I know it's turned all the time. But I just want to give context when we talk about customer experience and total experience, what we're talking about. The customer experience is that one we've highlighted. This is the Contact Center, our Zoom Virtual Agent, quality management, workforce management, Expert Assist, all those things that sit in that customer experience platform. In short, we are category complete to compete with a full breadth of product in the space. The other part we talk about total experience is how could I take my customer experience and connect them to field workers? I think there was that field worker app we launched today, like retail. If you think about a retail store or a grocery store, they have field workers and agents. How do you connect them? It connects to Workvivo.
When you start having these conversations with people that really want to differentiate brand, they're like all in. I mean, we're not creating this market segment. We just happen to have the technology to finally take advantage of what people have been asking for. So when we focus on upmarket, there's a few things we have to do from a product perspective. First is you have to have a reliable platform with privacy at mind and reliability at mind. I always say it doesn't really matter how many features you have. If it's not secure and it's not reliable, it doesn't matter, right? Especially in customer, you hear all the press. If Delta Airlines, you couldn't get a hold of a travel agent when there is a weather condition and people are trying to book travel, it'll be all over the news. It cannot go down, right?
So we just launched five nines SLA, which is very interesting because we have a competitor in the space that has a name very similar to that that does not offer that SLA. And it's ironic, but we are offering it now. We do have PCI as well, which is critical. That's a payment card. So we launched that this year. And then also FedRAMP, which really allows us to play in the kind of a federal space for keeping data and access to the data and privacy controls all within the U.S. So we can get to very, very, very large Contact Centers. And we do have a Fed-focused sales team going after that space right now. Excellent. All right. I apologize for drinking. It's just water. Promise.
So the other one, when we talk about moving up market, it's not just stability, but you also have to develop products that, and it's really got an AI focus. So we talk about AI in two ways. One is the virtual agent, which is self-service, really, is what we're talking about. And we have three announcements coming out now. One is the multi-intent detection. So your first-generation bot was very much like, ask a question, get an answer. Ask two questions, you're getting one answer. Ask four questions, you're only getting one answer, right? It couldn't pick up that there were multiple things happening in that question. Now with multi-intent detection, you can have a very natural conversation. Hey, I want to change my address and I want to buy something. And it'll say, "Great." And it'll take you through a script on how to solve both of those.
Very powerful stuff. We also have an integration to the history of the consumer. So we know not just what's happening in that engagement, but all the engagements that customers had with your brand over time. So we can make it more personalized. We know their buying behavior. We know what they've called about before. And we talk about having natural conversation, human-like conversation. This is what lets us to have it, make that happen. And then the other one is we are launching voicebot. So all that kind of intelligence that we're doing in chat and messaging and SMS and WhatsApp, all our digital channels, we're going to make available on our voice channel. And it's a critical differentiator because if you had a different vendor doing your voicebots and yourself, and someone asked a question about there's all these different vendors doing different things, right?
I had a voicebot over here and I built it, and this is how I handle the interaction. And it's a different vendor. And then I have a different chatbot, right? And this is how I handle my chatbots. It's different technologies. They're using RAG differently, if at all. They may not have multi-intent detection. And as a consumer, if I call in and say, ask a question, that experience will be different on the voicebot than it will be on the chatbot. Different answers, different tone, different experiences. And the one thing brands don't want to do is provide different experiences based upon the modality you contact them. So this is one of the benefits of that. But AI is not just about virtual agent. It's also about how do we make the agent's life better.
There are some people that say with AI, agents will die overnight and there will be no agents evermore. That is not true. I don't believe that. I've been doing this forever. That happened when speech analytics came out, agents are going away. No, they didn't. When auto QM came out, supervisors are going to go away. No, they didn't. There will be a decline, but they're not going to go away for sure. So that means we have to make sure that those agents are better prepared. So we have dynamic agent guides. So if there's a certain workflow I need to get done, guide me. Like step one, get their address. Step two, confirm their billing, their ZIP code. Step three, it'll just tell the agent what to do based upon the interaction type.
And if the customer had already given me their address and that's in the workflow, it's not going to guide the agent to do that again because it knows you already have it, right? So it's going to guide you dynamically based on what's happening on the interaction. Suggested answer, again, so you can have agents focus on engagements, personalized connections. It'll just say, "Do this, ask them this, tell them this, offer them that." So the agent doesn't have to think about going to knowledge bases or looking up processes and procedures. It just tells them what to do. And then also for supervisors, alerting them. So if there's something happening that they need to be aware of in a bad engagement, alert them automatically. They don't have to go and find it. So the other area we talk about focusing on is partner-led strategy.
A lot of logos on this slide. There will be a test. This is mostly to show, like we understand when we talk about going to market with channels, not all channel partners are the same. So we go to like VARs, value-added resellers, and these are your typical Cisco partners and Avaya partners that have relationships. They work with us in a certain way. GSIs, like PwC, that does large transformation projects, we work with them in a different way. Service providers that provide voice for our customers that have relationships with those accounts and want to do it, we work with them differently. So we just know that when we say go to market, this is a discipline in and of itself that we figured out, and then we also have co-sell, a recent announcement where you can buy Contact Center on the AWS Marketplace.
AWS sellers get credit for it. We get credit for it. Everybody's happy except for the Amazon Connect seller, but everybody else is happy. So in terms of like some basic stats on how we're doing, 2,500 accreditations so far. That's people have gone and been accredited on some aspect of the ZCX platform, selling, SE design, implementation. This is how you get your amplify, your reach, and get more into more accounts. 63% of our bookings in the first half of this year are through the channel. And when I say through the channel, I mean they source a deal. We didn't know about it until they called us. So we actually expect that number to grow, especially going to international markets. Some markets like Japan, which is a prime market for us in our global market, is 100% channel. So I expect that number to go up.
Massive growth, 165% YoY growth in deals coming through our channel network. When we talk about partners as well, we're not talking just about sell-through, sell-to, sell-with partners. It's also about our ISV network of partners. We do this for two reasons. First is this gives us access to verticals or market segments that we wouldn't have natively. SpinSci, NovelVox is an example. They have very good integrations with Epic and healthcare or Jack Henry and FinServ. So by partnering with them, we get into better positioning for different verticals. Or we want to move up market. We need PCI Pal to move up market to do payment processing.
The other thing it allows us to do is when we go into a Contact Center opportunity, it's very rare that they're going to say, "I'm going to replace everything I have in my Contact Center day one, and I'm going to replace it with you." So we have to integrate with some existing infrastructure. Salesforce will always exist in probably 60% of our customers, right? We need to integrate with Salesforce. They might have a Verint WFM or a NICE WFM. We can integrate with those platforms. You don't have to pull that piece of it out. So it's very important to view it as like, I always say, we love our suite of products and we think it can do everything for you. But if you want to integrate with other products, we are a platform as well and we can do that.
So that's kind of our partner strategy. The next slide will surprise all of us. There we go. This is about the, in terms of with these focus areas, what kind of success are we having? So 43% increase in our average per seat pricing. This is half one this year versus half one last year. This is because we brought so many new products to bear. We brought WFM, we brought QM, Expert Assist. This is driving up our average per seat price just in the first half versus the second half. And this is, we haven't even gone to like the second half of this year, but as we bring more product, I expect those numbers to get bigger and bigger and bigger in terms of the growth. 35% of what we're selling today, so the first half of this year is our advanced products, Premium and Elite.
So they're getting AI Expert Assist. They're getting our WFM. They are getting our QM. And this is something we just launched, I think, in January of this year. So launching something in January and having it become 35% of your sales in six months is, I haven't seen that happen before. And then 12% of our sales are coming from ZVA, self-service, and that number will continue to grow because self-service is top of mind for everybody. So we are seeing successes in the areas that we focus on. Thank you. All right. So like my mom always says, the proof of the pudding is in the eating. So I want to take you through a few case studies on where we're actually winning, and then we can wrap it up. But the first one is a really, really large U.K. council out of the U.K., or clearly.
But I bring this up for a couple of reasons. One is because it shows we are growing not just in the U.S., but in our international markets. And the second thing, this company actually had a CCaaS vendor in place. So this was not first-time move, premise to cloud. It's a vendor you all know or heard of, probably Genesys. And the initial approach was, "I just want self-service. I have CCaaS. I don't need anything else. I have a phone system, but I don't have self-service yet." So they were talking to us about just our ZVA. Now that that's bad, right? We'll sell ZVA as a standalone all day long, but that's what we're approached. And then we start talking about, "How about elevating from ZVA seamlessly to an agent, live agent?
And how about transferring that call from a live agent to a back office worker?" The specific use case for them was, this is the public calling in to kind of their, not their emergency services, but the county service. One of the services they provide was mental healthcare for the public. And if someone called in and they were having, and they determined that person needed mental healthcare, they would have to give that person a number and say, "Call this number," which is if someone needs mental health, that's probably not what they're going to do next, right? So they would seamlessly transfer over and make that a very connected total experience platform. This is total experience coming to life and saving lives in some cases. Next slide. This one I call out for a couple of reasons. One, it was because over 1,000 agents.
We competed with some of the usual suspects. It was a Cisco incumbent. The beauty here that I'm highlighting is this came to us from a business partner. We did not know about this deal. They called us and this business partner had a relationship with this brand for 20 years, and the brand went to the partner and said, "We need to move to a cloud Contact Center. We have a very good relationship with this partner," and they called us and we had one other competitor in there through another partner and we ended up winning this in large. They wanted the AI-first approach. They trusted Zoom. They didn't have Zoom. Here's the thing. They bought Zoom for the first time for our Contact Center. They didn't buy Zoom because they had already invested in us, which is a lot of our business.
And they ended up buying WFM and QM and everything else from us. So this is an up-market deal, partner-led. They're doing the services. And frankly, it truncated the average sales cycle in this deal by more than half because the selling had started with a partner in that brand months before we even got engaged. This is the beauty of selling through partners. And then the last one, also an international deal. This is one of the top grocery retailers out of Australia. And they were all a Cisco house, Cisco UC, Cisco Contact Center, Cisco everything. And they wanted also to connect not only the Contact Center, but they had 800, they do have 850 stores across Australia. And there was no way to connect agents to stores, to our customer calling into a store back to an agent. It was so disconnected.
But they told us, actually, "We're just looking for Contact Center. Don't talk to us about Phone." Okay, so what do we do as good sellers? We start talking to them about Phone because there was a solve there, and they quickly saw the value in it, and then we ended up selling both into that account there. For the last one, just put some numbers on this because what we do all the time is we don't sell technology for technology's sake. We sell it for outcomes, which is the next slide here. This is a case study, Chris Crosby, who owns a BPO, business process outsourcer, which if you know the industry, you know what it is. If you don't, it's like a very nuanced, but they provide agents for other brands.
As opposed to if I had a brand and I didn't want to provide agents, I would go to a BPO like Influx, and they would provide the agents and the technology to provide my customers' customer service. That's what they do for a lot of brands. He went all in on Zoom with everything, ZVA, QM, voice, Docs, the agent assist, Expert Assist, everything he went in. The outcomes were phenomenal. You look at this reduced after-call work from four minutes and 30 seconds to just 30 seconds. What I mean by that is agents didn't have to take notes after call. Here's what happened. Here's what they were looking for. By doing kind of after-call work, massive reduction in spend. Reduced the average handle time by three minutes because they're guiding the agent.
So the agent doesn't have to go to knowledge bases and search anymore, right? And then also the overhead, like from a technology perspective, people hate managing multiple systems. Like if you go to most IT leaders, what is their number one strategic objectives? Like I need to reduce my tech stack, not add to it. So he reduced like six different products by getting, and it started with the Contact Center, but it went into docs, email, everything else. He said he cut my IT admin for my technology stack in my company 50%. Huge savings. So I thin k that's the last slide in terms of just numbers, mostly go-to-market focus, but again, it's like a focus strategy we're working on.
The market is huge, and we are positioned not because Zoom's creating a market segment, but because the market is demanding CX and EX convergence and AI, and there's no vendor that can do what we're doing right now. So thank you.
I hope there's no questions. I think we do have time for probably five, 10 minutes for a few questions. That's okay. All right. Well.
Awesome. Yeah. Will Power with Baird. Chris, yeah, thanks so much. That was a great, great overview. Maybe just a quick kind of two-parter. You know, given the focus of market, are there still kind of key features that Zoom needs to add to be effective?
And then number two, just given your long experience, track record in the industry, including eight years at NICE, just love to kind of understand what you think really is going to set Zoom apart, particularly as you move up market, what really attracted you to Zoom and what's going to enable you to win.
Yeah, no, that's a great question. So our competitors will say Zoom doesn't have the features and aren't ready to move up market because that's what they want everyone to believe. I've been doing this for 25 years, and it's not true. Now, there is an exception like WFM, which is a kind of a beast in and of itself, and there's probably like two WFM real enterprise vendors, and it's Verint and NICE. It's not Zoom. It's not Genesys. It's not RingCentral, right? It's not in there.
But that's okay because when Genesys and Five9 and all these other vendors go and go up market, they're partnered with NICE and Verint too in WFM. So our gap is their gap because it's like the WFM stuff. But outside of that, no, there's, I mean, you might get a nuanced integration that it's not that we can't do it. We just can't say, "Hey, we have three deployments of this yet," right? That's where we get stuck, but it's not like there's something you just can't do as much as that. And then in terms of the disruption, like with CCaaS, it's really unsexy to talk about the platform because there's not like everyone talks about outcomes and AI and Expert Assist and self-service. But where these, like the first-generation cloud vendors got stuck is their architecture wasn't as enterprise-ready as their story was enterprise-ready.
You get into a sale, they love the story, you get it, you start talking about architecture, and it just didn't scale. It wasn't as reliable as they needed. These big brands like banks, major banks, major healthcare companies, even if they love your story, but you don't have the tech stack, they're very, very, very cautious to make that move. Zoom's tech stack from an architecture perspective is way better than theirs. I mean, like you talk to technical people, they're like, "Wow." They're like the things like scalability, reliability, failover across multiple zones in Amazon, all these things that the other companies can't do, we have in spades. It's that the fact that we have the architecture where they were kind of getting stuck in the up-market deals, we've done, right? We're ahead in that space.
So that's very exciting because I know where I got stuck in the past, and it was painful, but I don't have that anymore. I don't know if that answered your question, but.
Yeah, thanks, Matt VanVliet from BTIG. I guess maybe dovetailing on the end of that question there, but how much of this is customer education, willingness to push a lot of these interactions into a more automated fashion versus the technology finally catching up to maybe where the industry wanted to be a number of years ago? And then how much of an advantage does Zoom have bringing an entirely new platform to the approach and not having all this legacy tech debt that doesn't allow some of these more modern features to be in there? So kind of three buckets, but where do you think that shakes out today?
Yeah, so I mentioned at the beginning that the AI, when I say AI, I'm talking about AI both for self-service and AI for making agents' lives better, both of them, is the fastest adopted technology in the Contact Center space I've ever seen before. So a lot of this is, in some cases, you have like this FOMO from CIOs like, "I need to get AI in here." They don't really know where or how to use it, but they're asking for it. So there was more of a pull in that AI conversation than some other technology we've offered in the past from kind of the AI perspective. The other thing you mentioned that even these cloud vendors, they have so much technical debt. I would argue that it's 50% + still on-premise or private cloud product.
Now they call it cloud revenue because they shove a server in the cloud and now it's a cloud product, right? But it's still very technical debt in terms of it's just like a hosted private cloud stuff. It's harder to like when you start designing product and you're like, "Well, how am I going to fix this for the premise stuff and the real on-prem and the hosted prem and my new cloud platform?" That's hard. So what you're doing is you're crafting a story that sounds really good, but on Tuesday it works here and on Thursdays it might work here because of the architecture. Zoom has none of that. Everything's cloud, new cloud, right? So it allows us to innovate way faster in features.
And you talk to anyone like customers, analysts on how fast we've innovated, it's because we don't have to worry about all that technical debt that these other vendors have to. And they get stuck in there, right? They get stuck in some of these large complex deals. So I don't know if that answered your question, but that's.
Unfortunately, I know you guys want more, Chris, but unfortunately, we have our executive panel here now and we want to keep them on schedule.
Come on. I was on a roll. I'm kidding.
Thank you, Chris. All right. Thanks, everyone. Just welcome up our panel. And just as a reminder, if you have a question, raise your hand, I'll bring the mic to you.
And if you are in the virtual audience and you want to ask a question, please submit one via the Q&A function on Zoom and we'll try to get to those as well. Welcome, everyone. All right, Kash.
Hey, I'm Kash Rangan, Goldman Sachs. Congrats on the conference. Eric, one for you. You're now getting into domains that are not just focused on collaboration. You're getting into customer experience, employee experience. Decision makers, when it comes to those kinds of technologies, are very different than the ones that you typically sell to. So how do you make this transition from a company perspective to being seen as a broader collaboration platform that touches multiple stakeholders and constituents, front office, back office, and not just the professional frontline worker that's looking to do a call on Zoom? Thank you.
Thank you, Kash. You always. Yes, thank you, Kash.
You always have great questions. I thought I have a former CFO here, new CFO here. My life would be much easier. I don't need to answer to any questions. And fortunately, I still needed to, but that's a great question. First of all, I think it's not that straightforward and easy. That's the reason why many years ago when we decided we are going to focus on departmental applications, take essentially all kinds of business services, marketing department, we already have physical events. And for sales department, we introduced the Zoom Revenue Accelerator. Now we double down on Contact Center, which is focused on supporting teams, very different buyers. We know that challenge. We know that problem. But you look at our vision, Zoom already evolved into an AI-first work platform. When it comes to work, it's not only just about collaboration and communication anymore.
How do we help those department use cases? That's becoming more and more important, and this is the reason why, first of all, we understand the challenge. And then internally, as Chris shared, and we have lots of changes in terms of go-to-market strategy. That's the reason why we double down on our channel strategy. It used to be look at our meeting service, by and large, driven by our internal direct sales team. Now we double down, triple down on channel, different pricing, the strategy as well. And also we also needed to make sure we have a better product. That's a key, right? Because the business services like a Contact Center and a core collaboration product, they can seamlessly integrate together. Essentially, back to the transition, we nee d to make sure a better product internally ready in terms of the go-to-market side.
And I think that down the road, as long as our patient is really showing the value, win the trust, I think it should be okay. Again, it takes some time, but look at what we're doing on Contact Center. We're making very good progress.
Yeah. Oh, great. Thank you, Pat Walravens from Citizens JMP. Michelle, this one's for you. Yeah, I know. Did you sign off on this on the new long-term? Because if not, we're going to have to go through all this again in a few weeks.
Sorry, the question is, did I sign off? I didn't sign off on the long-term model that we shared this morning. Yes, we talked about it yesterday before we get that. I'm only here to set her up for success.
Okay, so you're good with it?
Yes.
All right.
And then the follow-up is just, I mean, obviously, coming from Microsoft, we all want to know why you took this role, right? And what you think that your experience is going to help bring to Zoom.
Yeah. So I guess maybe to start with why I came from Microsoft, I think it starts with a deep appreciation for what Zoom was in the pandemic to the world, respect for an iconic brand. And having been on the other side, a real admiration for something that sounds so easy, but it's maybe not, which is just being rooted every single day in the customer love and wanting to show up and be reliable and innovative every day for the customer. And just, I don't know, it sounds so simple that you would do that. And yet, I think to many places, it's not necessarily intuitive.
So it kind of started for me on that. But then when I started to look at Zoom 2.0 and where we're going with that and so many of the things that I think you heard in the keynote and here in our time together, so much expansive TAM, whether it's talking about bringing meetings to new customers and frontline worker or moving into natural adjacencies, even as Kash's question and employee experience or Contact Center, I think it just speaks to how much Zoom has really grown from just their stronghold in meetings. And so that's super exciting for me. I think customers want sort of that holistic point of view. And then you add in the AI, and I think maybe that's been the delight.
I didn't necessarily see this in the interview, but just hearing it today, the differentiated approach to AI, just having listened maybe on the other side to customers, it gets confusing with a bunch of copilots. They're not always sure about business value and paying for it. And so it's just an exciting three days in here to really hear the approach. And I'm still learning the federated and all the smart cost structure behind that, but I'm excited about that, obviously, too.
So you start doing what it notices with AI, you can learn everything quickly.
There you go.
Thank you. Finally, just quickly, right? So a few years ago, right, when I met Kelly, I think Kelly, I really want to grow our revenue to a billion-dollar revenue. Kelly said, "Yeah, I want to join. I can help for sure." Now I talked to Michelle.
We really want to grow our revenue to a billion-dollar revenue. Michelle mentioned, "Yes, I can help." Yeah, same thing.
I don't know, maybe I didn't answer the second part, but what do I think I can bring? And it's just having seen growth and a ton of transformation over 25 years. It was interestingly the CFO of Teams in the pandemic. So a lot of familiarity with the space and also having just seen a ton of transformation and growth. And so both on the engineering side and on the go-to-market. So I'm excited to be here.
Thank you. Samad Samana from Jefferies. Appreciate you having us here today. It's great to see everybody. Kelly, maybe one for you. The decline in the COGS, I think, was really interesting, right? And the average AI user and the costs related to that.
So my question was, how much of that cost decline per user is the price dropping that what your third-party vendors are charging you, right? So the third-party platforms versus growth in the number of users driving the cost per user down versus the colo efforts. And I guess as you look forward, where do you see the biggest levers and continue to drive that lower over time?
It's a combination of move to the colos, getting all of that traffic into the colos as quickly as possible. And then the growth, we had some fixed costs, of course, associated with AI Companion. And so now as we're seeing more activated users, that economy of scale is really kicking in.
Also, by the way, next time we might invite our President of Engineering Product, Velchamy, here. That guy's a genius.
He knew how to manage the COGS and very, very smart, very creative. Every day he thinks about, "I can save money here, move money to AI, a lot of things." It's an amazing leader here.
Peter Levine with Evercore ISI. Maybe one for Michelle. You're at Microsoft. And if you think about what Zoom is trying to do in terms of owning the Workplace, if you think about what Microsoft is doing, what do you think the shortfalls were with Microsoft that you think Zoom can take advantage of? Meaning you saw pricing strategy. What do you plan on doing now at Zoom, given what you've seen on the other side? I know it's early, but to the extent that you could share any initial thoughts.
I don't know that I'm going to share the details of it, but I think, look, both companies have their strengths, their pros, their cons. And I don't know, I would kind of go to focusing on what I think Zoom does brilliantly, the things that I talked about, the customer love. There's some things I think to work on and go-to-market and how we kind of approach that. So maybe there's some lessons and scale there. But rather than sort of make it this versus that, I think there's a lot just to focus on about Zoom and its differentiated approach. And a lot of just what makes Zoom wonderful, which again, I kind of go back to that product love and customer-centric kind of mentality. So I'm sure there'll be more details as we go, but.
Great. Thank you, Charles. Michael Funk from Bank of America.
Kelly, we had a clarification on the slide earlier with the margin guidance. I want to make sure I heard you correctly. I know you said that revenue growth acceleration is still the primary point of emphasis for the company. Did you then say, though, later that if revenue growth accelerates, the margin might be lower?
Well, we've said if we see opportunities to drive revenue acceleration, either organically or inorganically, we would potentially bring margins down to accomplish that. So if there's an investment that's needed to accomplish top-line growth, there potentially could be a decrease in margins over a period of time. We will always continue to strive to be as efficient as possible. But investing for top-line growth is the number one priority.
Sure. And so tying that back, then the long-term margin target, how does that link back to a revenue growth projection for acceleration?
Yeah, we aren't going to comment specifically on the model today other than just to say we had a very outdated framework, then we wanted to move towards something that we think reflects more where the company is today. But again, we had 39% margins in Q2, so we're being very disciplined and focusing on our profitability, but also want to keep reminding everyone that we will do everything we can to re-accelerate growth. And that's what the long-term margin model represents.
Great. Thank you. Yeah, honestly, honest impression here.
Sure. Hey, guys. Thanks. Alex Zukin from Wolfe Research. I'm going to ask a relatively long question to do my best impression. For actually for you and for Michelle. I joked with Kelly coming out of the keynote today that my favorite new announcement was monetization of AI Companion 2.0.
It seems like a lot of enterprise software companies, they have kind of a growth disease right now. They're all growing slower. It seems like there's a potential AI cure for that out there. So my question is for you first, how do we think about monetization, growth, re-acceleration in your mind? Are we at this kind of new point where all of the front-loaded, amazing kind of growth from the pandemic that we've been digesting and indigesting to some extent, we're getting past that, and now we have this new ability to really get further? Then, Michelle, you were CFO of Teams during the pandemic. That was a really important monetization tool and timeline for that product and the company.
When you think about what opportunities you see for bundling, for pricing, for packaging, and the changes you can kind of help make, what stands out to you? So kind of a two-parter.
Michelle, you want to address the second part?
You want me to address the first part?
You go first.
Okay. I think, Alex, unfortunately, I do not have a crystal ball, right? I do not know. And we already got past that or not. But I can tell you for sure, and we are getting closer and closer towards that goal. You look at over the past few months, you do see a few companies, they announced a layoff, right? Over the past 12, 18 months, a lot of companies, you see less and less, right? That's a very good sign.
As we look at our monetization in terms of driving growth, we cannot just look at one aspect or one service, one area. We need to look at everything, right? First of all, AI, like customization, AI Studio, Companion 2.0, and AI monetize our Contact Center or other services, right? I think this is a very good way for us to grow rapidly. That's one way. Another thing, how to introduce new services. We announced the Zoom Workplace for Frontline Workers, right? And for healthcare use case, also double down on vertical use case. That's another one, right? And also don't forget about we not only focus on the collaboration and communication services, but also we have multiple business services as well, right? And that's what we have today. How to have more? We acquired Workvivo, essentially focused on also different use cases.
We want to make sure we expand more and more new use cases, not only just for horizontal collaboration, communication use case. That's another way for us to drive growth. Either we build organically internally or we can go through the M&A as well. So we look at all those ideas, solutions to help us drive growth. But the bottom line is we want to make sure we know for sure there's no economic recession. And this is good news. Plus, and you see, as I mentioned earlier, number of layoffs is much less, and that's also good for us. So essentially take some time for us sort of get to a very solid foundation, we can grow. I do not know next Monday, next week, next month, but we are getting closer and closer. So that's a very good news for us. That's why we're very excited. So yeah.
Maybe I'll just add on with a comment about AI monetization, not necessarily from my prior experience or my job two days in, but I think the Zoom strategy is a great one, right, about democratization of AI. And then monetization when things get in finding that right balance. I think just my observation of customer reality is there's a lot of hesitation. Customers have been through so many tech transformations that they want proof and quick proof this time that AI really adds value. And so I think there's a lot of really smart stuff. I love the focus on Zoom 2.0 on really tangible actions and results. Again, I think it just gets back to this kind of customer-centric. That's the result of it. And then packaging, I'm not going to comment on prior, and I don't know enough to comment on my current two days in.
But I guess I would maybe just share two thoughts on bundling, which is I think it's always done best when the customer is kind of at the center of the concept. What are natural adjacencies of things that make sense from a customer perspective? Or maybe Kash, sort of similar to your question about audiences, right? I think at least from my prior experience, it's those natural adjacencies of an audience where you're selling to an audience or you're selling product where there's adjacencies where I think bundling makes more sense.
Hi, thanks. This is Kylie here for Tyler Radke at Citi. I'm just curious, is there a goal with the Custom AI companion for industries to drive more of the free-to-pay conversion? Is it that these industries maybe are more likely to have users on the free version of Zoom that would want to upgrade?
Any penetration stats with frontline or education or healthcare would be helpful. Thank you.
So yeah, I think we shared some of the data on the customers that we have on healthcare and education as part of the keynote. We shared a lot of data. So we talked about 90,000 schools and universities using 140,000 healthcare providers using us as well as our share in the telehealth market. So we do have Zoom Workplace for Education, Zoom Workplace for Healthcare offerings. And we are actually going to be giving them, they get AI Companion included as well. And we are expanding on those features to give them even more value. We think that helps us be even more valuable solutions to these verticals. So that's important.
But on top of that, if they want to really customize and add data sources, which could be any kind of other classroom records or lecture materials coming or even third-party providers or work across apps, pulling data from EHRs or LTI systems. So that is where we are offering the custom add-ons. So to answer your question specifically, I think our goal is to continue to keep providing more value and engagement and where it makes sense for them to, yes, to monetize with these customization options.
Hey, Michael Berg from Wells Fargo on. Maybe one for Eric or Smita. You guys are in an interesting place in the overall ecosystem. You have APIs that connect to every major application out there. And today, it seems like on the AI side that you are primarily doing Copilot type of work.
Maybe you can help me understand how you think about agentic AI and how that could fit in with the broader Zoom story and moving forward from here. Thank you.
I think maybe we did not communicate well. You look at the AI Companion, essentially that's an agent, right? Because we want to have a digital assistant, right? So AI Companion panel, right? Will help us get all those work done, organized. That's essentially indeed based on the agentic technology, right? I think that's, it's not like a Copilot style. And probably after you get that Zoom AI Companion, if you had a first-hand experience, I think you will feel like that. That's indeed our core innovation, right? How to leverage agentic technology to improve our AI Companion. That's kind of our vision. And maybe we did not communicate well. So yeah.
Hey, Meta Marshall, Morgan Stanley.
Maybe the flip side of Kash and Alex's question of just, Kelly, you gave sales or sales and marketing targets that are kind of reduced from your old targets. However, you're selling a lot of new products into a lot of different kind of customer types. So I'm going from kind of a core Zoom sale to Workvivo and frontline and some of these vertical specifics. So just kind of what changes are taking place from a sales force standpoint and how are you kind of accommodating those more efficiently?
So as a reminder, for Contact Center, we have a specialist team that is overlay. The Workvivo team is largely still acting as a business unit. And both of those teams, those specialty teams, partner with our core AEs. And then the AEs themselves will get trained.
We'll have a whole enablement process around Zoom AI Companion 2.0 as well as the verticalization. As a reminder, we do have a healthcare and an education team. So they'll obviously be the specialists in those areas. So I think it's a combination of having the specialties that are really deeply versed in their product areas and having enablement across the broader sales force. And then layering in, as Chris was talking about, really continued focus on the channel, which is needed especially to help drive the strategic growth trajectory in Phone as well as Contact Center.
Matt VanVliet from BTIG. Maybe digging in a little bit more on the channel side. There's been kind of start and stop on the channel strategy over the last few years, especially as the packaging around the entire platform has changed.
But obviously, the addition of Contact Center, they tend to be more of a gatekeeper there than maybe in other areas of the business. So maybe on kind of a two-part question, one on the growth side, how much is the channel also going to have the ability to sell the rest of the platform and maybe even get credit for renewals of meetings with the expansion of Phone and Contact Center or something more of that nature? And then second, on the margin side of that, how much of a channel mix maybe is contemplated in the new operating structure? How is sort of the marginal margin on new revenue via channel partner instead of a direct sales approach?
So in terms of the model with the partners, this has transformed, of course, over time because we had so many customers that were direct meetings customers.
But as we expand, we want to have the partners be able to sell Phone and Contact Center with them. We are actually even working on an internal upgrade to our systems to enable a lot of this and make it easier for us to work with the partners in those situations. And then in terms of the margin, even as we've seen fairly significant growth in the channel contribution over the last couple of years, we've continued to be very disciplined about our margins and been able to absorb that into our margin structure that you see today. So we will continue to do that. I think Michelle sort of hinted. I think one of the areas she's going to want to focus on pretty quickly is sales productivity.
So continuously trying to self-fund, if you will, any increase in potential channel expenses by driving efficiencies and productivity through the rest of the organization.
Thank you. Hi, Samad from Jefferies agai n. So just to follow up on the AI monetization and the prior question. So we're also seeing different pricing models in the market when it comes to AI. You can do a consumption model. You can do per query pricing. You can do per conversation. So just curious in terms of what were the drivers in moving to a licensing model like you have. And the second is, what about the margin impact versus a CSP, like an AWS or whatnot? Are you going to be paying higher margins to the model providers and the cloud providers for AI-related stuff, or is it going to be similar to corporate margins right now? Thanks.
I don't know if I can speak to the margin side, but in terms of how the pricing was determined, and we did announce that $12 per user per month for the Custom AI Companion add-on for Zoom Workplace, we looked at the features that we talked about. We tried to project what the usage might be. We estimated the costs for it. And of course, at this point, I mean, these are estimates, but based on the knowledge that we see we have from other features and the way we see the distribution across the customers, that's how we came up with the pricing based on usage. And then we also have a federated model in the backend, and we continue to optimize the technology stack as well. So those are the drivers that went into determining the pricing.
So, the margin impact we've modeled, and you heard all the reasons this morning, federated approach, as Eric said. Our DevOps team is amazing. And so, given all of that, even with what we are forecasting in terms of accelerated uptake and usage of the product, we expect the margins to be right in line with the long-term model we announced this morning.
I'll go back with a second one. Kash Rangan, Goldman Sachs. So, clearly, AI has become. Generative AI has become more accessible to companies, software companies that have the data and the ability to build models on top. Whereas a couple of years ago, it was not that clear that the incumbents have so much of an advantage. Typically, tech cycles are disruptive to incumbents, and we have new companies that come on the scene, right?
So in this case, you have the advantage, but so does everybody else. Salesforce has got their agents, and Microsoft does, ServiceNow does, Intuit does. So all of a sudden, this is becoming a free-for-all for everybody. Everybody has access to the same LLMs and Anthropic as you do. And you have your own company-specific model as well. So you have this federated thing. So who really has the advantage here? Is it the company that's got the control point for selling the technology or the use of the technology or the company that's got the best technology stack? Because it's not clear to me anybody really has any real advantage here, but I'm sure you think very differently, and I'm here to learn.
I think a great question. And I did a talk about this with many CEOs. That's a very wonderful question. We always talk about this.
I think two things. First of all, just to look at Zoom, right? Before AI, Zoom was just a communication company. That's what they say. No content, no data. After you and Casey and I had a meeting, Zoom is done. That's it. With AI, we are going to transition our business from a communication company to a work platform. Meaning, let's say after the you and I have a meeting, after meeting Zoom, guess what? We can generate a transcription. We can generate a summary. And you look at a summary, some of them will be the tasks, the action items, right? That's the reason why we introduce Zoom Tasks. Or we can integrate it with a ServiceNow, right? We just announced that the joint partnership, right? You can become a ServiceNow ticket, right? So essentially, every meeting itself will generate a lot of content.
It becomes a system of action company, right? It's very important. Finally, we can own a lot of data now because of AI. Without AI, everyone, we need to hire somebody to write notes, right? That don't work. That's very important for us. Other companies are different. They already generate a lot of data, much better position. Because of AI, finally, we are, I think, on the same category where we also can generate tons of data. That's really good for us. So back to the second part of your question, I think it's early for any new technology, right? For now, you look at probably at this one who makes more money, more profit. I think it's more like a, you look at a technology stack, right? It's kind of the chip company or the cloud API company, right? Those two benefit a lot.
That's more like on the first phase of any new technology adoption. But download all those applications. Unfortunately, for now, look at the entire industry. There's no killer app yet. But eventually, there will have more killer apps. But that's the future. But for now, I think the winner will be wherever can leverage AI, add more value to customers, to improve the customer interaction, improve the customer experience, whoever is going to win big in the future. Otherwise, let's say we double down on AI company, look at some of our competitors. They do not have those key features. And also, they charge a customer. Guess what? Those customers are smart. They are going to move to us, right? And download actually, maybe we can introduce some killer either features or apps, right? It's more like an early adoption phase.
That's the reason why I think we got to be patient as long as we keep investing, double down on AI, make sure to look at everything from a customer perspective. I think every company will be fine. The key is you need to make sure to have a very, very long-term sustainable strategy to embrace AI. If you today say, AI, I really think about let's monetize, charge a customer $30 per user per month. Guess what? You talk to many CIOs, ask them, "What do you think about that technology from our competitors?" I do not want to mention the name. You know that. What's the value? You talk to many CIOs, they say, "I really do not see the value." Just yesterday, I talked to the CTO. He's a very big bank.
I asked him, "What do you think about that technology?" Again, I don't want to mention the name. See, I stopped. We do not allow any of our employees to use that technology. It's no value. So yeah.
Yeah. Hi, this is Andy Kurita with Kettle Hill Capital. I had one question for Smith. I was wondering, you have 300 million sessions, daily average users. How can you use those sessions as a marketing tool to introduce the elements of your AI strategy to the people that already use your product for a different purpose? And then I had a question for Eric as to how he, the keys, the fundamentals behind creating a happy culture for thousands of people that work at Zoom.
Yeah, so I actually don't know how many sessions we have, if it's 300 million or not.
But we do have a lot of users who are on Zoom, and they are meeting on Zoom, and they are engaging with us. And I do believe in product marketing is some of the best ways in order to educate those users. We can put out a lot of marketing materials, but users don't really read them. They're using the product. And while they're using the product, if we can communicate the value or additional features, then it's a good way to engage with them. So I think those are some of the things which we continue to experiment with. We do have a product framework. We try to show them some we will put out a message like AI Companion 2.0 is here. If you're not using it, use it.
If they end a meeting, we are trying to thoughtfully think about how we can communicate some of this value to them to tell them that you can continue the meeting chat, or you could have generated, used AI in order to do more things over here. So we are looking at how we can continue to communicate more value to the user. The balance which we have to find is that we are not spammy towards those users, and we do not come in the way of them doing their critical job. So we do it in ways that's useful. So there's a set of those kind of features and how we expose them to users and how we help them turn it on then and there. For example, for AI Companion, for if you're a really small business, you're likely the administrator as well.
So can you turn it on then and there when we tell you about it? So those are some of the things which we are looking at. And for free users, our online team, they also run experiments in giving free trials and seeing how those convert.
Yeah, so back to the culture question. I can tell you, I think prior to the COVID-19 crisis, I think Zoom had the happiest culture. Every morning when I woke up, I was smiling. I said, "Wow, this is indeed the company I dreamed working for." But for now, I can tell you the culture is still broken, but for sure, much better than last year. And it takes some time to fix that broken culture. The reason why we made a huge mistake, we hired more than 6,000 people during the COVID-19 crisis, essentially less than 24 months.
When you hire 6,000 people, how can you have a good culture? It does not work. Because the reason why when it comes to culture, you need to build a trust, right? Not to mention a lot of employees work remotely, and we do not know them, and so many employees, and to build a trust, it takes some time, and I never thought about, I say, I talk about a digital twin. Our internal employees challenged me. "Oh, Eric, you do not understand AI. Why are you talking about digital twin?" I told our employees back in the office twice a week. Employees challenged me. "Wow, we are a remotely working company. Why do you want to employ back in the office?" Right? And I pushed the team hard to deliver something earlier, employees also challenged because they do not trust. They do not trust leadership.
They do not trust me. And that's a culture broken because of that trust relationship, because of 6,000 people. I think if you ask me where we are now, we are way better than last year. Are we out of the woods yet? Not yet. But we are getting also closer. And probably another few quarters, we are getting there. Every quarter, we look at the employee workforce, what can do differently. And someone very loyal, someone long-term employee, double down on them. Someone say, "Hey, you have a different opinion. That's fine. Maybe it's time for you to leave." Find some other opportunities. I think that's the number one priority for us. That's the number one thing that keeps him up in the night. Again, no shortcut. So yeah. Yeah. Because of our mistakes.
We'll take one question from the virtual audience.
We have Arti Vula from J.P. Morgan on behalf of Mark Murphy. He's asking, "You mentioned that frontline workers potentially constitute 80% of the workforce, but only 1% of SaaS spend. In that vein, how would you characterize the customer willingness to spend on the frontline versus knowledge workers? Are customers poised to increase spending on these sorts of users, or is there some degree of customer education that needs to happen first?"
So yeah, so as we built our frontline worker offering, we actually talked to a lot of customers. We talked to several customers in retail. We talked to customers in healthcare, and we talked to customers in manufacturing. So we talked to these customers in order to design the product and to understand what their needs were. And actually, these customers were asking us for a solution for their frontline workers.
The other thing which I would say is that Workvivo is really loved by a lot of companies that have frontline workers. It's deployed wall to wall. And with Workvivo, they're able to get engagement because now their employees are connected to the corporate news. They can see how the communication flows. But they were still missing tools for on-shift communication management. And they were asking us, Zoom, like, "You have strengths in these areas. Create an offering for us over here." So it was developed very interactively with our analysts as well as our customers asking us for solutions in this space.
Maybe I'll just add on with sort of my prior experience, which is I think there's a lot of conventional wisdom out there that you can't monetize frontline workers.
And I think it's really about finding maybe just what you're talking about, that customer value, and then you unlock it. Quite often, for many customers, that is the closest thing to their revenue or an area of their workforce where they're struggling to have retention. And so I think it's really just finding that good balance and mix.
Great. Yeah, Will Power with Baird. I think you all indicated at the keynote, and I think Kelly may have had in her presentation as well, that 57% of the Fortune 500 have enabled AI Companion. It'd be great to get any kind of framework for what the active usage looks within that. I assume it's some subset of that.
But depending on what you're able to provide there, maybe just any other color on kind of the trends you're seeing within the Fortune 500 on usage and just kind of the learnings that could help you get up to that, that folks that have enabled it really using it more actively.
So I think with generative AI, customers rightfully have a lot of concerns and fears. So the enablement journey itself is a long journey. So the way we have a very customer-first focus. So we actually, with Zoom AI Companion, we went out with a very clear posture. We don't use company customer data for training our model or third-party models. And we didn't put any ifs and buts around it. It's not that if you give us consent or if you do this. So we took a very clear posture.
We also provide a lot of granular controls. Even so, as we engaged with them, and almost so many customers use Zoom, which have very high compliance requirements. The Fortune 500 companies obviously have a very high degree. As we engaged with them, they also wanted to do a lot of due diligence, privacy, security. How are you? How's the data going to flow? What kind of controls do I have? So it's through that journey that this enablement has happened. So this enablement itself is a huge step forward, and then from the enablement, the usage is then expanding from there. If you're enterprise customers, you would start with a test group, and then you would think about departmental policies. If you're finance or if you're legal, you will have different policies, and we provide all of those levels of controls.
So that's the journey that we are seeing is that deep discussions, enablement, test groups, and expanding from there. You also heard from Praniti, who's the CIO of Zscaler. Their companies are getting to the point that they've given all of AI Companion capabilities to all their customers, all the employees.
By the way, just that I got a quote from a customer yesterday talking about how they are using Zoom AI Companion. Maybe a little bit long. I don't want to read it, the whole thing. So every day we got a lot of quotes like this. So yeah.
Charles, I think we have time for one more.
Yeah. Right. Hey, guys. I'm double do the double dip on the question side.
So I think to the former question, tying two threads together of does including more sophisticated 2.0 functionality actually lengthen sales cycles and slow down monetization from a time perspective? And then the second question I asked this, I think when you guys launched Zoom Phone or Zoom Contact Centers, if we start to think about the attach and compare it to previous new products that you've launched, whether it's Phone or Contact Center in terms of percentage of, I guess, in this case, seats within customers that would adopt it, not just customers, because it sounds like it would be a staggered journey. How do we frame those things as we start to think through the future?
So yeah, so for the AI Companion 2.0, when we think of AI Companion 2.0, we also think of it as a feature, if you will, where you get a side panel.
It's contextual. It can synthesize data across Zoom data, but also you can connect your Microsoft Outlook or Google email and calendar, upload documents, go to the web. So all of this functionality is available to the customers as incremental administrative controls, which they can turn on or off. So we expect this to be a progressive journey. The first thing is, have they enabled AI Companion? Oftentimes, they'll start with things like meeting summary, meeting questions, and they will progress from there. So I don't think it, I mean, we will see how it plays out. I don't think it elongates the enablement cycle. But yes, we do expect they will assess how much they want to turn on for AI Companion 2.0. The other thing which I would also say is that our customers range from really small to really large.
We have a substantial number of businesses that are solopreneurs or what we call really small businesses with under 10 employees. So we see different levels of just adoption and ability to use the features. Some of the, as you can imagine, the largest companies have a lot of concerns, and smaller ones too, but there's a variety.
I mean, this will really be up to Michelle, but I think that as the team starts working on FY26, these are newly launched products, right? So they're not contemplated in the growth or the, we think about them, and we are talking about from a margin structure, but from a growth perspective, they'll be taken into consideration as they work on FY26 modeling.
Maybe one more over there.
All right. One last question. It's going to go to Pat, and it's either really good or really bad.
I'm sorry if I missed it. Should we take that 33%-36% operating margin? Is that how we should be thinking about next year?
We are not giving FY26 guidance today. But no, that is a long-term margin structure, just the way you said, to give us a framework in which to think about how we would invest in long-term potential in inorganic or organic growth. But Michelle will be on the next call to give Q4 guidance, and then in Q4, we'll give FY26.
Okay. All right. Thank you, everyone, for joining us for our investor session at Zoomtopia. I have a few announcements for the in-person folks, just logistics where we're going next. But I think we can say so long to the virtual folks. Thank you so much for joining us, and look forward to seeing everyone next year.
Thank you, Charles. Thank you both.