ServiceNow, Inc. (NOW) Earnings Call Transcript & Summary

Good morning, good afternoon, good evening, and welcome to this webinar. Today, we're going to cover AI and generative AI in the context of service operations. My name is Paul van Nistelrooij, and I'm representing the strategic engagement team of our platform BU. And my role is to enable our customers on our strategy and vision when it comes to platform and generative AI capabilities. With me today are Aaron Zuber and RJ Jainendra. RJ, may I ask you to introduce yourself and then hand over to Aaron.

Thanks, Paul. I'm RJ Jainendra. I'm the General Manager for ITSM, and I lead the product team here for ITSM at ServiceNow. It's really great to be here to talk to you about Gen AI and its impact on service operations. Aaron, over to you.

Thanks, RJ, and Paul, for the introduction. It's nice to be here with everyone. My name is Aaron Zuber, I'm the Global Area Vice President for Service Operations Sales and thrilled to share some insights and some knowledge around generative AI and how we're using that within our service operations framework. Back to you, Paul.

Thanks, Aaron. Thanks, RJ. Let's kick this off, the first section. So the agenda for today's webinar. We will start up with covering AI and generative AI and what it means to your business and to ServiceNow. And then we'll focus on how AI and generative AI accelerate service operations. Throughout the webinar, you can post your questions in the Q&A panel, and we're happy to come back to those and see if we can answer those at the end of this session. So let's kick off with the first section. And before we kick off, I would like to learn from you and where you are in your journey when it comes to generative AI. So specifically to generative AI or gen AI, where are you? Where are you in your journey? Have you started? Have you not started yet? Are you in a research modus? Are you trialing some solutions? Or are you in the process of implementing solutions? Or have you already implemented solutions? So if you can please pick your choice here. And while you're doing that, I would like to reiterate the option to post questions in the Q&A panel. Also, there's a resource section at the bottom of your screen where you can find useful links to more information on these topics. I can see that we have a number of almost 20% of the attendees, we're moving along. So please pick which is most relevant to you. Have you not started your gen AI journey? Are you researching? trialing any solutions? Doesn't have to be ServiceNow, there's other solutions as well. Or are you in the process of implementing gen AI solutions in your processes already. Just north of 30%. Really hoping we can get a bit more. So if you can maybe think about them and see which option best reflects where you are in your journey, that would give us a bit of an idea of where we are. Maybe 5 more seconds. We're almost at 40%. That will be a nice benchmark to reach. Okay. I think we're close enough. I'll close the poll and let's see where we are. Let's look at that. So almost 50% are in the stage of research and solutions and 1 in 3 have not yet started. In fact, all these answers are good for this webinar, right? So you're all in the right webinar. So we're going to talk about some of the background, we're going to talk about some of the benefits of generative AI within the platform of ServiceNow and specifically to the context of service operation. So lets get moving. Right. If you look at today's world of digitization, we strive to serve our customers through digital services. And what we see is that getting things done is not always easy. Left hand side, you'll see that requester, you'll see a number of channels. And a lot of the digitization is pushing those requests to various channels. At the same time, companies struggle to meet customer expectations because disparate processes and technologies are being tied together by human in middleware. There's always humans involved and there's always a challenge with getting everything aligned. And just last week, I had a pretty poor process experience when I left my phone in a plane after a business trip, which is not a wise thing to start with. But even at the airport, they were kind enough to find my phone and put a picture on their website. And when I claim the phone to be mine, I had to give proof, which I did. After I claimed the phone, it disappeared from the website without giving me a notification. So for almost 2 days, I was unsure whether my claim was acknowledged or not. What made it worse is there was no way to contact the loss-and-found department. How do I e-mail, phone, chat, whatever. And with all the best intentions of the people at the airport, the clunkiness of their process led to a poor end user experience. What made it all worthwhile is I got my phone back in the end so was happy nonetheless. So, this can be traced back to agents in middle and back office to manually handle tasks and communicate the different departments in separate systems. Now if we look at the loss-and-found experience that I have, this is really an example of a digitized process or partly digitized one. And this is what we see in every business around the globe. An ever increasing number of services being pushed digitally with the intent to better serve employees and customers. And at the same time, an increased number of monitoring and alluding tools generate an increasing number of events and alerts leading to machine-generated incidents. The combination of these 2 results in an exponential increase, the number of incidents, both human and machine generated. Now in the old days, you could scale these processes by just simply adding more resources to it. With this exponential increase in incident volume, it's pretty much impossible to scale people and manual processes to keep up. This is why you need AI-powered automation. To process these incidents as they come in or maybe even proactively address the issues before they generate incidents. Aaron will elaborate later on more on the view what essentially depicts how we see service operations operate. Before going there, let's have a look at how ServiceNow has invested in AI and ML capabilities over the year -- over the last years. So if you look at the last 7 years, you can see that AI is not new to ServiceNow. We've been investing in this area for a long time, and we've been growing the platform organically through talent acquisitions as shown in this overview. We started in 2017 with our virtual agent. We introduced the machine learning platform. We add an extra Natural Language Query and understanding capabilities through that and so on. The one acquisition I want to call out specifically in the context of generative AI is the acquisition of Element AI, which was our first step into the world of enterprise and generative AI. And with this acquisition, our journey in what now is generally known as gen AI, started. Here's where we laid the foundation for Now Assist product suite, which we will cover in the next slides. Before we do, lets have a look at how customers are realizing value from our in-platform machine learning and artificial intelligence capabilities already. So here's just a selection of some of the customers that have embarked with us on that AI journey. Veolia is a global water, waste and energy management company who uses performance analytics to establish a best practice approach to KPIs and performance measurement, resulting in a 75% improvement on service response times. American University provided a consistent, personalized 24/7 experience through the use of virtual agent. Accenture automated the issue assigned to workflow using in-platform machine learning capabilities to increase the speed and accuracy of the ticket assignment to support over 0.5 million employees and a ticket volume of over 10,000 a day. And Sanford Health increased employee productivity through the use of virtual agent and machine learning ticket assignment on transferring cases between departments. Bottom line is, AI has been an essential part of the Now Platform with a lot of our customers already. All right. So let's move on. So AI is only as powerful as the platform is built on, right? Let's zoom in on the hype. Every vendor out there is racing to deliver gen AI capabilities. And most vendors, they are offering integrations to general large language models like OpenAI. This is where the Now Platform and ecosystem are unique. Our approach is to build the main specific LLM straight into the platform. That means that we've abstracted the complexity of dealing with AI models. And in doing so, we can offer gen AI capabilities to support their processes and easy to consume model. Around the workflows you've already invested in, driving the action across your enterprise. So because we build gen AI into our platform, is a platform feature, all of our workflows are going to be able to benefit from gen AI and that's why we can innovate in this space faster than anyone else. Let's talk about how gen AI is such a game changer. Previous forms of AI knew how to identify patterns to train models, generative AI has the ability to create new patterns in 3 steps. First to understand is context, so understand the intent of a request or an issue or whatever communication comes in. It also understands the sentiment of the person making that request. Secondly, it's able to go through multiple information sources and combine them into a relevant understanding. And then 3, and this is really the magical step, this is where we generate content in a way that effectively communicates the findings from step 2. So if we look at it in real life for example, if a customer asked when will my refund be processed? Gen AI understands that the person, obviously this is not a positive, right, it's probably a bit of a negative sentiment here, but also it understands the intent. This customer is looking for information. The Now Platform then searches across multiple sources. This can be sourced from inside and outside of ServiceNow platform, knowledge bases, ERPs, whatever. And it's finding the relevant information. And then step 3, assisting the customers interacting with your brand via virtual agent. They will receive an accurate and concise answer with the relevant information. For example, as soon as the return merchandise arrives at a warehouse your refund will be processed. This slide really shows our approach to general purpose and domain specific large language models. And in ServiceNow, we rely on those 2 pillars when it comes to providing the best gen AI experience. On the one hand, we will leverage the general purpose large language models or LLMs, where conversational aspects are relevant. For example, through dialogues, through our conversational interfaces. This is what large language models like OpenAI, Azure and Google Bard are really good at. In order to provide a better experience when it comes to ServiceNow out-of-the-box use cases, we built our own -- as an example, if you want to use something like text-to-quote where you generate scripts as coding based on natural text. If you would use OpenAI for that, you'll get a result that looks remarkable and remarkably useful. But an actual fact, it will not work from a ServiceNow platform as we have our own specific version of Java code. And Now LLMs are specifically trained to model to understand these specific characteristics so that generated code will work. And that is one example. Other examples I think of understanding what resolution notes are? when to generate those? If you create a knowledge article, what approval process to kick off, et cetera, is really understanding the context of ServiceNow and leveraging that knowledge to provide a better experience. So how will generative AI transform the enterprise? So will generative AI change the AI landscape? It's a fascinating technology, and I think we all acknowledge the enormous power, the potential power that it has. When I talk to customers, they all want to know, what does that mean for me? What does it mean for our organization, for our processes for our people, for different personas in the organization. Also how can ServiceNow help? So I start off by stating that things done in today's digital world is not always easy because of the interaction between digital processes and human intervention. With ServiceNow, we put people first. Our customers create and orchestrate the digital workflows on our platform. And the platform already helps them connect every system in digital process across their enterprise. And when we inject these processes on our platform with AI and generative AI, we will transform the experience for people working in your mid and back office, your employees and customers, creating a massive productivity catalyst that will add enormous value and create enormous agility within your organization. All right. So let's hear it from one of our ServiceNow customers. I'm going to play a short video, and after the video, Aaron will take over from me and zoom in how we -- you and I to bringing AI and gen AI together and unite that with our service operations vision. Here we go. [Presentation]

Well, I love that video, and thank you so much, Paul, for that introduction and just laying the foundation for how important and how impactful AI specifically generative AI can be. I wanted to talk a little bit about uniting services and operations. The goal being is how are we going to do more with less. Before I jump in, though, I do want to say that the slide that Paul showed with all of ServiceNow's acquisitions of the different tech that together comprehensively represents our AI and generative AI capabilities. It's important to know if you're not already familiar with it. When ServiceNow acquires a technology, it's folded into the platform and we replatform it, meaning that you as leaders and practitioners of the technology do not have to maintain separate skill sets and have this hub and spoke setup going on with it. So it provides a very, very seamless experience and it drives the narrative of our native platform, which is really the value add with ServiceNow as we start to explore these different technologies and continue to add them to the portfolio. So I wanted to talk a little bit about Digital-first business growth and what happens when it's disrupted by technology services and operations because of the siloed nature of it. So, the teams, the tools and the data, when they are siloed within and among departments, it really makes it hard for them to provide services to employees and customers. And when we take a look at each one of these stakeholders and how it's negatively impacted by basically a proliferation of different disparate tools. It perpetuates lack of automation. So these teams that really should be working together are not and why is that happening? So IT service teams were not able to meet the increased number of requests from employees resulting in poor experiences. And I think Paul covered this briefly in the idea that there's this explosion of digital services. So we're going ahead and we're trying to -- we're trying to promote the different services that IT is trying to offer to the rest of the organization. But as a result, if we don't have that back end of the service supply chain ready to go, it can create a bottleneck upfront. And when IT operations teams are not able to predict an event service outages, this results in lost productivity and eventually revenue. Employees and customers are struggling to access the services they need to be productive. They can't self-serve, which is a big, big problem and actually a big desire from them. They want to be able to take care of themselves. They're waiting too long for available agents, and the result is that they're unhappy with IT. So there's a big perception and experience issue here. Employees, they're frustrated with IT support, tech teams are formed with business with no consistency and management of tech vendors and systems. And then IT operations, like I mentioned before, they're dealing with major delays resolving high-priority incidents. And the list goes on and on. And the result is there is a very, very poor experience. So the question really is, how can we leverage gen AI capabilities and the consistency that we have on the platform in order to improve these experiences, reduce the outages and the resolution times and actually get ahead of these and prevent things before they start to take place. So at this point, I really want to turn it over to the audience for a poll. How do your services and operations teams collaborate? Are they totally siloed, which is -- in a lot of cases, this is the way things are traditionally set up. Or do some have manual collaboration or some -- do some use the common services data solutions? Are there unified processes? So this is really -- this is -- if we can go ahead and answer in the A, B, C or D. What I'll do is I'll give you some time to answer. I would like to get up to as close to at least half the attendees if we can. We have a lot of -- we have a big crowd now. But again, are these things that -- there might be a smattering of both which is completely fair. So total siloed teams which -- I came from a practitioner background myself, I can tell you that in my experience, my teams were very much siloed, and the collaboration was a big struggle. Are there manual collaborations? So are there some interventions that you have that are still very manual? Do you use some common data and solutions or the last one, do you use unified processes, tools and data across teams? So that's -- would be really where we'd want to be ideally. But where are you in your journey? So I see right now we have over 42%. I'll give you a couple of more seconds to respond and then we'll close the poll. [Voting]

Okay. I think we have enough for the poll results. So Interesting. So almost 50% of you said there's a little bit of a mix here. So we have some common data and solutions. Roughly 1/3 of you said there's manual collaborations. Interesting to see that under 10% had totally siloed teams, so there's been a lot of progress here. And then 10% on the other side of this curve have unified processes, tools and data across teams. Very, very interesting here. So let's talk a little bit about AI-powered service operations, really what it is and what it can bring to the equation. So if I think of achieving digital-first business growth, it really starts by bringing technology services and operations together. And by bringing them together, you can expand technology services while reducing costs, we're all being asked to do more with less. And the idea here is that we're trying to drive extraordinary employee and customer experiences. At the same time, we really want to make sure that we're driving and accelerating innovation and productivity. And if we're -- if we're tripping over our own shoes trying to communicate or trying to collaborate or running into silos, that can really [indiscernible] innovation and productivity. So as a result, you get customers and employees that are happy with IT and services if we do this correctly. And according to EMA, 67% of the organizations were ServiceOps initiatives rate the impact as very positive and in a lot of cases, transformative, which indicates the high-value return and relative ease of these efforts. So what we're doing is we're helping the teams work more naturally and intuitively together. One of the things that I would like to say is that something that differentiates ServiceNow from other predictive AI Ops solutions out there is that -- and Paul covered this in some of his slides, we're learning what's normal. And then we're differentiating from something that's not normal. And we're making sure that, that data helps with the actions that our agents and operators are taking. So like -- what we like to say is when we're talking to a customer, we say, hey, the best incident is no incident at all. So how can we get ahead of these? So taking this one step further, and I'm just going to go ahead and build this out. We can take a look at a specific use case here. Danske Bank was a customer success story. And this is a number of -- I'm not going to read the entire slide here. But basically, what they're seeing is some significant reductions and significant return on investment, 93% reduction in high-priority incidents. So this was a massive impact to their business. They leveraged One platform for ITSM, for ITOM, so its service operations rollout for sure, service portfolio management, strategic portfolio management and governance, risk and compliance. So these were all of the things that they took advantage of from the ServiceNow platform. And they were dealing with a lot of things in the background, legacy tools built in silos, which is common to what we've seen here with the polls with a lot of you that are on the call. Inefficient patchwork of data processes and then ongoing customization that was difficult to maintain. So there's a little bit of a sprawl here and tech debt that they were dealing with. And the results they got were very, very impactful for them. And this is just one of the success stories that we wanted to highlight, but there is a lot more that we have ready to share with you. But before we do, I wanted to go ahead and turn it over to RJ to talk a little bit more about AI and reimagining services through generation or through generative AI.

Thanks, Aaron. So with the service operations, our goal is to help our customers deliver the best employee and customer experiences while running a robust technology operations center. And so AI based on machine learning and generative AI deliver a variety of outcomes that improve the employee experience, to improve agent productivity as well as prevent incidents and outages before they happen through AIOps. And so on the employee experience side, our conversational virtual assistant empowers employees with self-service. For common technology issues without having to get an IT agent involved, right? Virtual agent is available anytime, anywhere and employees can also get answers from Now Assist for search which takes the most relevant search result and runs it through generative AI to create a summarized answer with the most relevant information, saving them time from having to read through long documents to find what they need and really helps them improve productivity and is overall a better experience. On the operational side, we help accelerate service mapping by up to 7x with machine learning. With predicted AIOps, we apply AI to events, logs and metrics and so we can help customers predict and prevent issues before they impact the users of the business. So for example, we correlate incoming alerts from monitoring tools through machine learning to cut through the noise and identify the ones that are impactful and then leverage classification to route incidents to the right team to quickly resolve the issues. So there's not a lot of back and forth hot potato and the right team can get working on resolving the incident. Now with Now Assist for ITSM is our generative AI technology applied for ITSM use cases that helps you drive outcomes across employees, agents and service owners. And so with Now Assist, you can increase deflection rates with GenAI-infused search and virtual agent so that employees can help themselves through self-service. You speed resolution wrap times for agents with resolution of generation and lowered the overall MTTR for incidents, which is what every service owner wants. And so let me walk you through how we can reimagine the typical employee experience and also the agent productivity, leveraging Now Assist for ITSM. And so a typical workflow for an employee who has an issue is they may go to a portal and look for a knowledge article, they may search for an answer to a problem that they have. And if they don't find it, they may turn to a virtual agent to see if they can get the answer there. And if not, then there's typically a handover at that point from that virtual agent to a live agent. And as a live agent on the service desk comes on, they help to triage the conversation. They -- kind of best practices to read through any interaction that the user or employee has had and come up to speed. And at that point, they could start to solve working on the problem for the employee. And agents are trained to look for knowledge articles in case that same issue has happened in the past. And if there is a knowledge article, they can refer to it to help resolve the issue. And they also look for similar incidents for which maybe there isn't a knowledge article yet to see if there -- again, their resolution notes that might help understand what the solution may look like. Now if the level 1 Service Desk agent isn't able to address the incident, typically, they would then reassign those to an SME and the SME also has to then come up to speed, kind of look at the incident, what else has happened up to that point, and then work on resolving that incident. And at that point then typically, best practice would be to create a knowledge article so that in future if the incident occurs, then that knowledge article can be used to either drive that deflection if the employee can reference it or even for the service desk agent as well. So now let's kind of look at how Gen AI can help reimagine this process and make it a better experience with the employee as well as improve the productivity for agents. And so with chat summarization, that initial interaction that the employee has with virtual agent, that entire chat can be summarized so that as a service desk coming on and taking on that transition from virtual agent to the live agent, they don't have to go back and ask all the questions again of the end user. And I think we've all been on this, right, where we have to repeat ourselves over and over again as we engage with somebody new who is trying to help resolve a problem. And so not only does the agent save some time in terms of triaging the existing conversation, but it's also a better user experience. Similarly, if they have to escalate the incident to an SME, the SME can use incident summarization to, again, quickly get up to speed, understanding all the activity that's happened up to that point. When the incident is resolved through a resolution notes generation, we can help the agent quickly write what the resolution was. Oftentimes, when we talk to customers, we hear that -- because agents are really pressed to move on to the next incident, a lot of times, they may just write fixed or done for their resolution notes and move on. So by providing these tools to help agents write the resolution notes more efficiently. We can help them adhere to the best practice. And the resolution notes that are created do end up helping because in the future when there are similar incidents that occur agents can refer to this resolution notes to find the solution, which can also be used then to create knowledge articles, right? And these knowledge articles that are generated can then be used by agents as well as by the employees on the portal when they have a problem. And with the virtual agent, these knowledge articles can also be summarized to really provide the answers. So overall, it ends up being a better experience, both from an employee perspective and it also helps the agents in terms of their productivity and efficiency. So let me show you a demo of what some of those experiences would look like. So I'm going to switch here to show you our gen AI-powered virtual assistant and ITSM, where you can take your chatbot experience to the next level with significantly less investment and much faster time to value. So let me roll the tape here. Here we go. So I'm going to assume the role of Jason, one of our engineers here at ServiceNow. And I'm going to visit my team in India to collaborate on the next release. So I need to set up automatic replies when I'm out of office, but I don't know how. So what you're seeing here is the employee center portal where employees can search for answers and track with IT. And so over here, I can bring up the virtual agent in the employee portal and ask how to set up automatic replies. And so the virtual agent searches through the knowledge-based repository and finds the most relevant answer. And what you're seeing here is it generates an answer for me based on the KB article that it found and it gives me step-by-step instructions on how I can set up my automatic replies. Now when I'm traveling for work, I still need to handle customer escalations. And so I'm going to ask the virtual agent for an iPhone. So I'm going to come in here and say, "I want a blue iPhone 15 Pro with 256 gigabytes of storage and I need to buy tomorrow." Now the virtual agent understands my intent and matches it to a catalog item to order the iPhone. And over here, it looks like I can complete the entire process right here in the virtual agent, and that's great. So let's get started. I need to provide a justification. So I'm going to say that my manager has approved the iPhone for customer escalations. And I'm just going to be putting my existing numbers, I've declined the carrier service, and I'm not going to be returning my old phone. I'd go ahead and confirm that I am ordering the phone for myself. And at this point, I need to choose where I will have the phone delivered. So I go ahead and I pick the location where I want it to be delivered. And now I get a summary of all the options that I have chosen. Now the interesting thing here that I would point out to you is that the virtual agents understood that I wanted an iPhone Pro of blue color and the storage of 256 gigabytes, and it understood that right from that initial prompt and filled in those options. So I didn't have to answer those questions all over again. Now I can, at this point, go ahead and confirm my selections or make changes. So in this case, let me go ahead and make a change. And this is something that is really powerful with generative AI-powered virtual agent, where with previous technologies, I would have had to restart the entire flow and answer all those questions all over again. But with gen AI powered virtual agent, I could simply say, "Hey, I need a new phone, and I'm going to ask for a black phone with Max storage and black color and the Pro Max." And so what you'll see here is that it understood again that I wanted an iPhone of Pro Max type, the color being black and it interpreted that I want the largest phone as 512 gigabytes. And so this is the kind of intelligence that is very difficult to achieve with the older technologies for a virtual agent. And so in this case, I can go ahead and submit the request. I get a link to the request that's been created, so I can track its progress and go back to my work. And so what you've seen here is that Now Assist for virtual agent powered by generative AI with intent recognition, intelligent slot filling and that multi-turn conversation capability is a game changer to elevate the user experience. And additionally, it doesn't require customers to build all the topics and flows for their knowledge base and catalog items. With a few configuration steps, you can expose your knowledge base and your service catalog to enable self-service and benefit from all the deflection right away. So let's come back to the slide. And -- so here's a quote from a customer, Coursera. Coursera is an online learning platform founded in 2012 by Stanford University professors and offers a wide variety of courses from top universities and companies around the world, including Yale, Stanford, Google, IBM. And courses are available in a wide variety of formats, including video lectures, quizzes, assignments. And so they've been an ITSM customer and an early adopter for Now Assist for ITSM. Coursera deployed the chat summarization and resolution notes generation across all its agents, and they are seeing increased agent productivity, where agents save time by reading in summarization and also in writing resolution notes by using the generated resolution text. And so Darla Wolf's quote here, she's been a strong ServiceNow Champion has continued to partner with us and provide valuable input to improve our feature set, roadmap and our LLM quality. So with that, I'm going to hand back now to Aaron to take us through how you can get started with ServiceNow in the AI journey.

RJ, thank you for that demo. I think that really helped to paint a clear picture for folks on the call about what some of the capabilities are and how that could make a material impact in their business. So always great to see a demo of the product. And there's a few questions that we're going back and forth to the chat. What we're doing is we're just scratching the surface right now. So these capabilities and power of the platform that you're seeing today, is just the starting point of how we're going to continue to layer on these very, very powerful capabilities in the future with subsequent releases. And in the next section -- and then we'll take it home and go into a Q&A. I wanted to share how to get started on some of the options with your ServiceNow and generative AI journey. So to kind of take a step back. And one thing I wanted to say is that a lot of folks and especially folks that I'll talk to is they'll say, "Hey, look, we've got a lot of different investments going on over here. And we're trying to have these initiatives over here. In some cases, we do have board level funding or we're investing in a -- in some type of an automation team. But how do we really get started, where do we choose the start? Are there certain objectives that we want to focus on first to get the max amount of return on investment and how do we really focus with this." So what you're looking at here is -- and this, by no means, is exhaustive or details everything out. But we typically find that customers start by digitizing manual processes that may include connecting silos or by looking for automation opportunities, such as process mining when considering the return on investment. There was a question in the chat that I went back and forth with that we're talking about how do we really measure the benefit of this and performance analytics is a very, very powerful way of measuring and predicting and then you tie that with process optimization and the intelligence that we have in there to identify the bottlenecks and then automate. It's very, very, very powerful first step, and it enables you to get to that return on investment quickly. And what we've seen is that as customers mature up at scale, they're really focused on establishing that foundation, leveraging the CMDB, CSDM, the common data models and then really attacking those engaging experiences. So kind of like what RJ went through where you have sentiment and summary and being able to deflect those calls and those requests through a virtual agent, you're really hitting a different level of self-service and deflection. And then the next step of that would be to optimize. And you're really trying to get into insights, making smarter decisions and using that to build in that automation, so future requests are benefiting from the same type of intelligence that they've gathered from fielding requests that were similar. And what we're seeing over time is that 20% of requests are being deflected through self-service. And some of the outcomes that we've also seen are, can be up to 66% reduction in MTTR, which is very, very significant. And then we have savings in the millions of dollars when we're talking about automation and improved quality of service. So very, very significant savings and return on investment, but it's -- like I mentioned, it's just getting started. So some of the key questions that I suggest and that we've heard asked over time and time again is, this is the tech services operations transformation journey. So can you achieve business goals that are easier if service and operations teams were troubleshooting together more effectively? So what are we seeing on one side that could benefit the way in which we're resolving with historically disparate communication on the other side. Are we really providing a collaborative and an intelligent framework for this communication to go on? How can AI then take advantage? And how are we going to take advantage of these different capabilities to improve service delivery or performance? So if there's a service degradation notification that's going out, what is the response to this? How are we employing these different generative AI capabilities to make sure that resolution is quick? And then when we do have siloed tools impacted, what are some service outcomes and operational KPIs that we're targeting here? So these are just some of the questions that are very, very common as we're considering where to start and what to do within that first journey. And then one of the things that we like to take after this is just to say, hey, look, when we develop a gen AI strategy that drives AI or that drives growth and lowers costs, what are some of the immediate steps that we should be thinking about and taking advantage of. And so the first thing I always say is like let's schedule a discovery session. Every journey is different. There's a lot of similar elements. And we can take -- we can take a different approach for each account and each person, each account is different. You're facing different struggles, and we can tailor a solution or a tailor an approach for that. Like similar to what RJ did, we'd like to show you the art of the possible. So getting -- part of this is because we want to show what has already been done in what is available on the platform. But also what should we be planning for in the future. A number of you are already going into FY '25 or getting close to wrapping up FY '25 planning. So what are the things that are going to be available next year, 6 months, 12 months, 18 months from now, that I should be aware of and I should be planning for in order to make sure that my strategic initiatives and capabilities are aligned. And together, we'll build a business value assessment and make sure that we're very specific about the different types of costs that we're trying to take out of the equation, or make sure that the multipliers, the value multipliers are very, very clearly defined so that we have some targets that we could shoot for. So like I mentioned, not an exhaustive list, but some very solid steps that we would like to take to move forward in the future together. And with that, what I wanted to do is I have 5 things to share in terms of resources. I would highly encourage everybody to visit a Generative AI and Service Operations sections on servicenow.com. Similar to what you've seen today with RJ and he just scratched the surface on a lot of things, but -- and these are all linked out. It takes time to watch some demonstrations on this. There is a number of different ways that we've employed generative AI on the platform. We're just getting started, like I had mentioned, but there's a lot more to come. So watch the demo, get familiar with what we're intersecting with Now and what's to come. I'd also encourage you, we have a thriving ecosystem, join the communities. So Now Platform, ITSM and ITOM, there's -- a number of your peers who are already there. It's a wonderful way to collaborate, to share stories, wins, questions and get plugged into the community so you can benefit from that. I'd also like to make sure that folks are aware that we're going to do another webinar on March 6. So this one is going to be about how are we going to do more with less leveraging AI automation and Gen AI to maximize your productivity. So a lot of similar themes. We, of course, will be publishing these for an on-demand format as well. So if you're not able to attend the live sessions, check out the on-demand webinars as well. With that, I believe I'm passing it over to Suzanne for the Q&A. While Suzanne is coming online, I think and RJ, I know you're here. I know Paul is online, too. Do we want to -- hey, Suzanne. And we're having trouble hearing you, too. So maybe what we could do is give me the thumbs up if you want to keep going, otherwise, we can field these questions.

So hey, what I'm going to do is I'm just going to jump right in here. And Paul, I know you were jumping through a number of questions. Are there any ones that maybe you wanted to highlight that you answered or ones that you see popping up from folks?

Yes, I'll answer a few questions that came in. There's a few questions coming in now in the last 2, 3 minutes and a lot of that has to do with some of the last slides Aaron that you showed, how can we get started? What is required for that discovery session. So I'll leave those to you. Other questions that came in are things like do we need to be on the latest release? And the answer to that is you need to be on Vancouver, related to that, I mean, working with service operations workspace is a must. So those are some of the things that I've answered in the past few minutes. And maybe you can elaborate a little bit on how customers can get to the next step with our discovery option that you just called out. Who do they reach out to and how do we get started?

Yes, absolutely. Well, that's right in my wheelhouse. I'd love to talk about that. There's -- we have wonderful coverage with our sales force. So what we want to do is get you plugged in with the team that is focused and dedicated to your company and account. We can take everything from there and make sure that you have all the material, all the interaction and all the contacts that you need, and we'd be happy to set something up. We also have, like I mentioned before, with the communities and the call to action with joining the communities. A big part of this is our partner ecosystem. And our partner ecosystem is a very, very close side-by-side partner when we're having these conversations with customers and a number of our partners have dedicated service operations offerings that tie right in with our gen AI capability. So in a number of cases, what we like to do is, as part of that, those initial conversations when we're getting to know what targets, what initiatives, what strategies are important to you, we also want to make sure that we're engaging the right partners so that you have the right horsepower and talent set behind you to take it from conceptual to framing it out and then actually executing against the strategy with our partners. So that's kind of an overview of kind of getting started, what I recommend, and that's actually as a practitioner that I was a few years ago, the strategy that I used as well with the partners. R.J., I know that you were going back and forth. When you have questions on demos and capabilities, is there anything that you just had a few minutes to go through. Is there anything that you wanted to highlight that maybe you didn't have a chance to go through on that demo?

So on the demo, I think I covered what I wanted to, but what I'll add is that a lot of times do get questions on what is the impact and benefit that gen AI can bring? And how -- is there a way to quantify it? What are the stats? And so we rolled out these solutions into only at ServiceNow. So we have a very robust Now and Now program where we drink own champagne. And so our digital technology team, our IT team rolled out the Q&A -- the KB summarization in the portal, rolled out chat summarization, incident summarization, resolution notes for our agents, both agents that serve customers as well as in internal. And what we found over the course of the last 6 months or so is that on the employee side, we very quickly got up to roughly 20% plus on self-service improvement. And this is a matter of weeks. So really helping kind of that employee productivity and employee experience. And then on the agent side, we're seeing that roughly about 50% of agents' time is being saved on the routine tasks, right? And so this is both within the -- for customer cases, customer support as well as within the IT help desk. And again, so that's really helping improve the employee set as well as for our agents, we are also seeing reduced agent turnover. So -- it's still early days in terms of the number of use cases, we're just scratching the surface, but really promising results that we're seeing just from our deployment within ServiceNow.

Aaron, can I check and see, can you hear me better now?

Suzanne. Yes, yes, we can.

Great. And I wanted to just kind of transition from the questions and let people know that we've gotten so many questions. We will be summarizing and sending out an FAQ so that your questions really are answered. I wanted to move on because we're running out of time and just let you know that we invite you to attend K '24. It's an opportunity for you to hear from customers and also ask questions to them at 10 labs and meet other people to talk about how you solve your problems together. It's a wonderful community and that will be happening in Las Vegas in May. In closing, I want to say a big thank you to our esteemed panel, RJ, Aaron and Paul for their insights into AI and gen AI for service operations. As we've mentioned, we wanted to talk about the 3 different layers. So Paul talked about the intelligent platform. Aaron went in to discuss how unifying ITSM and ITOM into the service operations concept adds value. And finally, RJ really gave insight into generative AI and how that can move your organization forward. Again, this is just the start of our conversation. We really invite you to join us in March when we continue this conversation. In closing, I want to let you know that we will be sending out an e-mail, you'll get the recording, the slides and some additional resources in the next day or two. And please do the quick survey at the end that pops up. We really want your feedback, what you liked, what you didn't like, what you want to hear more of. We're here to provide information that you find useful. In closing, thank you for your time today. We really appreciate you. And with that, may you have an awesome rest of your day. Thank you.