AI integration for enterprise ft. CJ Desai of ServiceNow
00:00:00.000 | (bright music)
00:00:03.200 | This year at AI Ascent,
00:00:05.800 | we are doing a few things differently.
00:00:07.840 | One of those things is the what's next section
00:00:11.280 | that we talked about, these visionary ideas.
00:00:14.080 | Another is the what's now that Sonia has been talking about,
00:00:17.540 | about how are people taking things in AI
00:00:19.920 | and implementing them.
00:00:21.760 | As part of this, we don't just have
00:00:23.960 | the amazing foundation model CEOs like Sam,
00:00:27.680 | and Arthur, and Daniela.
00:00:29.800 | We've also brought in two exceptional leaders
00:00:33.280 | in the space of taking AI
00:00:35.440 | and making it enterprise-ready at scale.
00:00:38.000 | One of those is CJ.
00:00:39.800 | CJ has been the president and COO of ServiceNow
00:00:45.240 | for the past seven years.
00:00:48.420 | And in addition to ServiceNow being a Sequoia-backed company
00:00:52.340 | for over a decade,
00:00:54.080 | ServiceNow is in an exceptional business.
00:00:56.840 | It's the kind of business anyone in this room
00:00:58.360 | should aspire to be like.
00:00:59.820 | Pat, I believe sourced the business in 2012.
00:01:03.840 | Is that right?
00:01:04.680 | 2009.
00:01:07.000 | And what was the scale of the business then, Pat?
00:01:09.460 | 20 what?
00:01:11.500 | Million of ARR.
00:01:14.160 | Okay.
00:01:15.280 | Anyone have a guess as to where ServiceNow is now
00:01:17.980 | in terms of ARR?
00:01:18.820 | Andy, I feel like you know this.
00:01:22.880 | Billion?
00:01:26.100 | One more guess.
00:01:28.280 | 20 billion?
00:01:29.120 | Okay.
00:01:29.940 | Somewhere in that range is correct.
00:01:34.880 | So ServiceNow is the third largest
00:01:38.360 | software-as-a-service company in the world.
00:01:40.760 | It has a market cap of $155 billion.
00:01:45.180 | It is about to cross 10 billion of ARR.
00:01:49.340 | It's at 9.75 billion of ARR.
00:01:52.360 | Remarkably, when you took over as president and COO
00:01:57.000 | seven years ago, it was at a billion of ARR.
00:01:59.880 | So it's 10x growth.
00:02:01.800 | And it was 13 billion of market cap.
00:02:04.480 | So it's been a 12 plus X multiple from there.
00:02:08.020 | It's adding 600 million of ARR a quarter.
00:02:13.480 | So let that be something to aspire to.
00:02:16.160 | And I mentioned it's the third largest
00:02:19.200 | SaaS company in the world.
00:02:20.860 | Well, number one is growing at 11% a year.
00:02:25.000 | Number two is growing at 12% a year.
00:02:28.360 | And ServiceNow is growing at 26% a year.
00:02:31.900 | So we all are good at math that ranking
00:02:37.480 | doesn't last very long with that kind of growth.
00:02:39.440 | And CJ told me virtually all of this has been organic.
00:02:44.360 | They've made some acquisitions like Element AI.
00:02:46.520 | We'll talk about that in a little bit.
00:02:47.880 | They've been very ahead of the curve on AI.
00:02:50.240 | But really it's been organic.
00:02:53.640 | It's called ServiceNow,
00:02:55.400 | but it does not have services margins.
00:02:58.200 | This is an 82% gross margin business.
00:03:01.600 | The free cash flow is 30%.
00:03:04.680 | Operating income, 27%.
00:03:06.680 | This is a rule of 56 business on the rule of 40 terms.
00:03:11.440 | And since IPO, since Pat sourced it in '07,
00:03:15.480 | I don't know what kind of multiple it's had.
00:03:16.840 | What was the valuation then, Pat?
00:03:18.480 | What it would be invested?
00:03:22.400 | 260 million posts.
00:03:23.840 | That's pretty good from 260 million posts
00:03:25.560 | to 155 billion, nicely done.
00:03:28.200 | Since IPO, it's been up 42 times, 42X returns.
00:03:32.560 | So an exceptional story,
00:03:33.840 | and one that I think is both grounding
00:03:36.240 | and also aspirational for everyone in this room.
00:03:38.980 | One of the main reasons why we were so excited
00:03:43.200 | to have you come, CJ,
00:03:44.520 | is ServiceNow has been way ahead of the curve on AI.
00:03:47.680 | We first met actually in a conversation
00:03:49.440 | where you were talking about the AI vision with NVIDIA,
00:03:53.720 | with Jensen and the NVIDIA team.
00:03:55.840 | And we might have a clip actually,
00:03:57.280 | but just yesterday at the NVIDIA conference,
00:04:00.640 | ServiceNow was a main feature.
00:04:03.040 | The specific use cases on the NVIDIA platform.
00:04:05.240 | And at the previous conference, it comes up as a main user.
00:04:08.440 | My first question for you, CJ,
00:04:10.720 | is please tell us how you got here
00:04:13.240 | to this exceptional role as COO
00:04:14.960 | where you do much of the product work.
00:04:16.760 | And then please tell us how you've gotten
00:04:18.440 | the AI suite, what it is today,
00:04:20.580 | and how you've gotten the AI suite to where it is.
00:04:22.360 | Yeah, so first of all, thank you for inviting me.
00:04:26.120 | And he asked me a fun fact.
00:04:30.000 | And the fun fact is that I'm a failed standup comedian.
00:04:35.000 | So I'm constantly working on the material
00:04:37.600 | to try something out,
00:04:39.440 | very high self-deprecating humor
00:04:41.800 | that you will see from time to time.
00:04:43.600 | And at the highest level,
00:04:47.440 | I want to say that we are extremely,
00:04:50.840 | extremely grateful to Sequoia.
00:04:54.360 | And here is a simple story.
00:04:56.680 | The story is that before we were going public in 2012
00:05:01.360 | is when we went public with a modest market cap
00:05:04.360 | of $3.9 billion.
00:05:05.920 | Okay, that's when we went public in 2012.
00:05:08.320 | And it was a meh IPO.
00:05:11.400 | People were like, you know, Facebook was pretty good.
00:05:14.440 | Back then it was called Facebook, the same year.
00:05:16.700 | And Workday was another one.
00:05:18.920 | And we were a meh IPO.
00:05:20.440 | But before that,
00:05:21.840 | VMware made an offer to ServiceNow
00:05:28.000 | for single-digit billions,
00:05:29.760 | single-digit billions, below five.
00:05:32.440 | And Doug Leone and the Sequoia team
00:05:36.120 | convinced the management team,
00:05:39.400 | which would have been happy to take that offer
00:05:42.320 | because they're like, wow, we don't have to go public.
00:05:44.880 | We can be part of a great company like VMware at the time.
00:05:47.840 | And Doug Leone, to his credit,
00:05:50.780 | convinced not only the board,
00:05:52.560 | but also the management team that you can become big.
00:05:56.780 | So we are extremely grateful for Sequoia's coaching
00:06:01.460 | at that point in time.
00:06:03.380 | And for the $2 billion offer which we could have taken
00:06:08.220 | versus today's $155 billion in just a matter of 12 years,
00:06:14.740 | it's incredible.
00:06:15.920 | So thank you to Sequoia and the team
00:06:19.640 | for sourcing us and believing in us
00:06:22.260 | and allowing us to get here
00:06:25.280 | because it would have been a very easy thing
00:06:27.960 | for Doug to say he was on the board
00:06:30.000 | and the return would have been amazing
00:06:33.360 | for Sequoia at that point in time
00:06:35.680 | and we still walked away.
00:06:37.000 | - CJ texted me before and he said,
00:06:38.920 | "Excited to see you.
00:06:39.760 | "Can't wait to be part of the cult that is Sequoia."
00:06:42.080 | You're already in it now.
00:06:43.040 | - I am, I am, I am.
00:06:45.500 | But I just wanted to be thankful and grateful
00:06:48.500 | and you guys work with great people at Sequoia
00:06:51.380 | and I never take that lightly
00:06:53.500 | and on the path to our growth,
00:06:56.540 | we never forget our friends and supporters.
00:06:58.700 | So just wanted to start there, okay?
00:07:01.260 | Number two, on the scale,
00:07:03.140 | so I joined the company,
00:07:04.580 | so Frank Slootman hired me
00:07:06.340 | and Frank Slootman was also placed by Sequoia
00:07:11.260 | at ServiceNow.
00:07:12.660 | And Frank had two choices at the time,
00:07:15.020 | which is I think public knowledge,
00:07:16.860 | but Frank's data domain gets acquired
00:07:20.020 | and then he decided,
00:07:21.780 | and I make fun of Jess all the time on this topic,
00:07:24.860 | but he decides that he's gonna be a VC
00:07:28.340 | and he joined Greylock.
00:07:30.460 | And if you have met Frank Slootman,
00:07:31.780 | which I'm sure some of you have,
00:07:33.940 | he is not a Sand Hill guy.
00:07:35.780 | So he said, "No, I need to go back to operational CEO job."
00:07:40.820 | And it was Sequoia that convinced him.
00:07:43.260 | Frank had two offers, Palo Alto Network CEO
00:07:45.900 | or ServiceNow CEO and he picked ServiceNow.
00:07:48.860 | Okay, at that point in time.
00:07:50.060 | And that was also Sequoia was behind it.
00:07:52.340 | So when Frank hired me,
00:07:54.220 | we were doing billion plus in ARR
00:07:56.540 | and Frank said, "CJ, by 2020,
00:07:59.940 | "if we can get to four billion of ARR,
00:08:03.600 | "that'll be a massive W and let's go for it."
00:08:06.940 | And our founder who created the company
00:08:11.500 | close to at age of 50, Fred Luddy, who became bankrupt.
00:08:15.880 | So this is not a classic Stanford, Harvard story,
00:08:19.300 | but it's like he went to Indiana University
00:08:21.880 | and it's originally from Indiana.
00:08:23.620 | And he said, "I'm gonna create a platform
00:08:26.300 | "that can solve multiple use cases."
00:08:28.740 | And so we always knew that TAM was pretty much unlimited
00:08:33.860 | and the platform provided everything you need
00:08:37.380 | to create a product or multiple products.
00:08:40.840 | And so when I joined, we were one-ish billion.
00:08:45.420 | We had one large product and two or three small products.
00:08:49.100 | And since then, it has been brutal execution
00:08:53.780 | on which buyers we go out of.
00:08:56.020 | Like the simple question on when my product team comes
00:08:58.500 | and says, "Hey, we can create this great product."
00:09:01.740 | My first question is, "Who is the buyer?
00:09:04.060 | "Do we have access to that buyer?
00:09:05.720 | "Is that buyer next door or two doors down
00:09:08.100 | "or five doors down?
00:09:09.460 | "And will current buyer introduce us to that buyer
00:09:12.180 | "or those two buyers don't talk to each other?"
00:09:14.340 | Those kind of simple questions on who is the buyer,
00:09:17.700 | who are we competing against, okay?
00:09:21.260 | And even though this is very simple,
00:09:23.400 | the third one, what is the size of the price?
00:09:26.140 | That if we nail this use case,
00:09:29.480 | can we create a billion-dollar ARR product?
00:09:32.560 | Again, another product.
00:09:34.800 | So we have been doing that level of precise execution,
00:09:39.800 | and that's what has helped us on organic innovation.
00:09:43.340 | We haven't bought revenue.
00:09:44.640 | We are the only SaaS company that has not bought revenue
00:09:47.040 | on our path to 10.
00:09:48.260 | And we always buy amazing companies,
00:09:54.280 | which has great people,
00:09:55.720 | and then we make them work on our platform.
00:09:57.940 | But that's how we have scaled now from 2016 billion-plus
00:10:02.320 | to 2023 when we exited the December quarter.
00:10:07.320 | We already reached 10 billion of ACV.
00:10:10.080 | And then, of course, the revenue trails a little bit,
00:10:12.760 | and we guided for 10.75 billion growing at 21%.
00:10:17.200 | I mean, these are some of the numbers,
00:10:19.400 | but it has been,
00:10:20.720 | hey, you have an underlying platform that's cloud-based.
00:10:24.160 | What products do you create?
00:10:26.360 | What are you solving for?
00:10:27.800 | Who is the buyer?
00:10:29.460 | That brutal focus on who is the buyer.
00:10:31.960 | Do we have access to the buyer?
00:10:33.760 | And what is the size of the price?
00:10:35.600 | And without that, with my product team and engineering team,
00:10:39.320 | and I joined as a head of products and engineering.
00:10:41.420 | Frank hired me and left after that,
00:10:44.960 | but that's been the focus.
00:10:46.440 | And on AI, it's as simple as we had a fundamental belief
00:10:50.760 | from supervised machine learning and as AI evolved
00:10:53.720 | all the way to Gen AI today.
00:10:57.200 | We have been very focused on AI in service of our use cases,
00:11:02.200 | because if we can infuse AI in our use cases,
00:11:07.160 | it's a very easy conversation with a JPMorgan Chase
00:11:11.480 | or a Citibank or United States Army,
00:11:14.640 | that, hey, you are using service now for this use case,
00:11:17.440 | AI will help accelerate X or accelerate Y.
00:11:20.840 | And so we have been acquiring or gaining small teams
00:11:26.700 | that are AI experts at various stages all the way from 2016.
00:11:31.700 | 2017 was our first one.
00:11:34.040 | 2016, we started the journey.
00:11:36.200 | And then when we bought Element AI,
00:11:39.060 | they were trying to be the next Google of Canada.
00:11:42.080 | And they had somewhere between 170 to 180 engineers
00:11:48.120 | between PhDs, data scientists, and engineers.
00:11:51.680 | And they were in this amazing team,
00:11:53.840 | a lot of very well-known people.
00:11:56.900 | Yoshua Bengio, who won the Turing Award,
00:11:59.300 | was part of that team.
00:12:00.780 | There are people who have written some seminal paper
00:12:03.020 | on transformer model.
00:12:05.300 | That's the kind of team we got.
00:12:06.940 | And the call I got from Allen & Company was,
00:12:10.340 | hey, this is a great team.
00:12:11.960 | They have no revenue, zero.
00:12:14.180 | And they are trying to figure out
00:12:16.780 | what use case AI can be applied to.
00:12:19.280 | And this was during pandemic.
00:12:21.760 | And I went to my boss, our CEO, and I said,
00:12:24.340 | hey, man, these people don't have any revenue,
00:12:26.780 | but it's a great talent.
00:12:28.540 | And we need to spend some money.
00:12:30.260 | And to Bill's credit, he said, absolutely.
00:12:32.700 | If you believe this is a great talent, let's take them.
00:12:35.060 | And they showed me Chad GPT 1.0, 1.5, 2.0 demos
00:12:40.060 | in 2020, late and early 2021.
00:12:47.140 | And then when this whole thing just blew up in 2022,
00:12:51.100 | we exactly knew where we could apply LLMs to our use cases.
00:12:55.260 | And again, I don't know if it's a term, but SLMs.
00:12:58.820 | We are very use case specific LLMs
00:13:02.200 | that we apply in service now.
00:13:04.860 | And we started our monetization strategy in September.
00:13:08.380 | So that's the story.
00:13:09.540 | I know it was a simple question,
00:13:10.820 | but I had to give a pretty long answer
00:13:12.300 | because there's a lot of history to it.
00:13:14.740 | - Really incredible.
00:13:15.940 | And I didn't realize, frankly, until this conversation
00:13:18.940 | that you guys are the only SaaS business ever
00:13:21.140 | to cross the 10 billion ARR point fully organically.
00:13:24.920 | That's remarkable.
00:13:27.320 | And frankly, I've talked to a bunch of people in service now.
00:13:30.180 | You are so much of the product brain.
00:13:32.580 | It's such a pleasure to get to learn from you.
00:13:35.980 | And this question is on product.
00:13:37.680 | So you guys had a little bit of a headstart a few years,
00:13:42.940 | 'cause I know you and Bill had been talking about AI
00:13:44.700 | even ahead of the element acquisition.
00:13:47.140 | But then with the element acquisition,
00:13:48.820 | you got to think about
00:13:49.660 | how you're gonna integrate it into your product.
00:13:51.140 | Tell us about how you got up the curve
00:13:53.360 | and now how AI is in service now products,
00:13:55.980 | maybe a couple of examples.
00:13:57.940 | - Yeah, so I'll just take the recent example.
00:14:00.920 | We are a big fan of open source community
00:14:06.300 | when it comes to AI.
00:14:07.860 | Even on the NLU models, we worked with Stanford
00:14:12.180 | to figure out which libraries we can use,
00:14:14.060 | which was four or five years ago.
00:14:16.460 | But we are a big fan of open source community.
00:14:19.180 | And the team in Canada, working with Hugging Face,
00:14:24.180 | figured out for which use cases of service now
00:14:27.300 | you can apply AI.
00:14:28.620 | So then we said, okay, now listen,
00:14:31.240 | you talked about our gross margins.
00:14:34.760 | Our gross margins are 82%.
00:14:36.960 | And all of you run the companies,
00:14:39.620 | your profitability starts at your gross margin level.
00:14:43.020 | That's your first step or staircase.
00:14:45.060 | 82, then you add R&D cost, sales and marketing cost,
00:14:48.180 | G&A cost, and then you get to profitability.
00:14:50.860 | So we are world-class in terms of our gross margin at scale.
00:14:55.860 | So I don't have the luxury.
00:14:58.380 | It's a constraint-driven optimization problem
00:15:01.300 | that I don't have the luxury to say,
00:15:03.720 | I'm gonna run open AI everywhere in my farm, in our cloud,
00:15:07.740 | because we are 100% cloud company with 170 billion,
00:15:12.460 | who knows, 2 trillion parameters now, with 4.0.
00:15:15.780 | I don't have that luxury.
00:15:17.220 | So the constraint was, can I run smaller models faster
00:15:22.220 | with lower latency for service now use cases?
00:15:25.980 | And that was a constraint-driven innovation.
00:15:28.520 | And we partnered with Hugging Face,
00:15:30.040 | our science and research team,
00:15:31.900 | and we came up with the first model on text-to-code.
00:15:35.420 | And we are not trying to do text-to-code
00:15:37.060 | like GitHub Copilot with Java or anything.
00:15:39.580 | Our text-to-code was specifically service now code,
00:15:42.940 | how you configure service now.
00:15:44.800 | And that was our first breakthrough
00:15:47.220 | working with Hugging Face.
00:15:48.820 | And then, once we do that,
00:15:51.620 | and you talked about Jensen,
00:15:54.200 | he's a big fan of Canada.
00:15:58.460 | So when we acquired Element AI--
00:16:01.500 | - Me too, Canada.
00:16:02.780 | - Yes.
00:16:03.620 | So when we acquired Element AI,
00:16:08.180 | it was the first phone call he made,
00:16:10.420 | and said, "CJ, love the Canadian talent.
00:16:14.180 | "We should do more together."
00:16:15.660 | And that was in 2020,
00:16:17.540 | because his history with you, Toronto,
00:16:19.500 | and ImageNet, and all of those things.
00:16:21.240 | So what we did is,
00:16:23.540 | that Canada team working with Hugging Face,
00:16:26.180 | we figured out smaller models,
00:16:28.420 | a one-tenth the size of OpenAI.
00:16:31.380 | And I told Jensen, "Hey, man, dude,
00:16:33.880 | "you are constantly pushing the next rev, H-class,
00:16:38.240 | "plus plus, I need these to run on A-100s."
00:16:42.640 | And that's what we'll work for,
00:16:45.120 | because we are a public company.
00:16:46.360 | So you have 1% gross margin, de-sell,
00:16:50.960 | and the number of questions I get from investors,
00:16:53.560 | like you get from VCs all the time, are not fun.
00:16:56.760 | So, yes.
00:16:59.600 | So we basically said,
00:17:02.120 | "I want smaller model that can run on A-100s,
00:17:04.880 | "and I can replicate that in every cloud."
00:17:07.600 | He's still always trying to push me.
00:17:09.080 | He's a great salesman, even though he acts like he isn't.
00:17:12.160 | He's always trying to push for H,
00:17:14.120 | to say H is faster, more efficient, which he's right.
00:17:17.560 | But we wanted something to run on A.
00:17:19.600 | So the smaller model,
00:17:21.280 | smaller models is where we are going with use cases.
00:17:24.640 | - Well, I hear you're really gonna need Blackwell.
00:17:26.560 | - Yes, I know.
00:17:27.400 | - The highest margin model.
00:17:29.080 | - $30,000, or something?
00:17:30.240 | - Exactly.
00:17:31.200 | What's that to you?
00:17:32.040 | - Yeah.
00:17:33.080 | - Fabulous.
00:17:33.920 | I have one more question for CJ,
00:17:36.120 | and I'm giving you that heads up
00:17:37.760 | so that you come up with a couple questions top of mind.
00:17:40.480 | CJ, we've got a room of builders here,
00:17:42.880 | and they're building many consumer companies,
00:17:45.600 | but also many enterprise companies.
00:17:47.480 | Frankly, you're a dream customer
00:17:49.600 | for a lot of the companies in this room.
00:17:52.080 | What can you tell people building products in this room
00:17:55.000 | to help guide them towards being a great AI builder
00:17:58.720 | that ServiceNow might consider partnering with
00:18:01.320 | and being a customer of?
00:18:03.560 | - Yeah, so I said there are two places.
00:18:05.520 | So just, when you look at me, look at my forehead,
00:18:08.240 | and it's a $1 billion spend I have in my cloud and software.
00:18:11.880 | So if you wanna sell something to me,
00:18:13.840 | make it quick, and I'll buy, okay?
00:18:16.240 | But I spend $1 billion, no jokes,
00:18:18.960 | on cloud and software and on the infrastructure a year,
00:18:23.160 | and growing at 25% in line with our revenue.
00:18:26.920 | So that's how much I spend,
00:18:29.520 | so I can be a great customer of yours,
00:18:31.480 | or at least a prospect.
00:18:33.780 | In terms of what works is,
00:18:37.740 | if you understand ServiceNow,
00:18:41.100 | which if you go to our website,
00:18:42.740 | you will not understand what ServiceNow does,
00:18:45.300 | but if you understand ServiceNow,
00:18:49.980 | we do basically a lot of workflows,
00:18:53.780 | as in tasks that get orchestrated digitally
00:18:58.780 | in a certain sequence between human and machines.
00:19:02.260 | That's what we do, right?
00:19:03.940 | Because people at a large bank or customers tell me
00:19:08.940 | that they can get a Tesla faster
00:19:13.700 | than getting a PC from the bank or a Mac from the bank
00:19:16.980 | when they order something, right?
00:19:18.340 | I mean, that's the reality of large corporations,
00:19:22.220 | large governments, and so on.
00:19:24.020 | So when you request a PC at a bank, say,
00:19:28.900 | and banks try to be very efficient,
00:19:31.180 | the processes that it goes through,
00:19:32.980 | does it require two levels of approval?
00:19:34.780 | Some banks have, for Mac, four levels of approval.
00:19:37.860 | Then once those approvals are done,
00:19:39.340 | it needs to go to shipping department.
00:19:40.740 | Do they have inventory?
00:19:41.820 | They need to base image it,
00:19:43.420 | they need to put security creds on it,
00:19:45.460 | and then it goes to, okay,
00:19:46.700 | what is going to her home address,
00:19:48.620 | or it's going to, these are the workflows,
00:19:51.460 | and these are the things that we automate behind the scene.
00:19:54.540 | So all you say is, I want to pick this PC,
00:19:57.340 | I want to pick this monitor,
00:19:58.860 | I want it to be delivered via FedEx tomorrow morning.
00:20:02.140 | That's the idea, right?
00:20:03.460 | But because of these complex workflows,
00:20:06.220 | and the banks want to harden the image of the Mac
00:20:09.540 | that they give you, if you're doing on a trading floor,
00:20:12.260 | that's what it takes.
00:20:13.100 | So that's what ServiceNow does,
00:20:14.380 | and then we infuse AI in making it simpler and faster.
00:20:18.300 | So for us, if you understand ServiceNow,
00:20:22.020 | and you can say, hey, CJ, for your use cases,
00:20:25.980 | here is the great technology that we have built,
00:20:29.520 | and you can consume this technology,
00:20:33.500 | whether it's your LLMs,
00:20:35.020 | or whether it's a use case specific AI that you have done,
00:20:38.380 | or some kind of analytic, whatever it is,
00:20:40.580 | then you have my attention.
00:20:43.780 | That if I can make the use cases for our customers better,
00:20:48.040 | I have your attention, and I'll buy your product,
00:20:52.380 | to make us go faster,
00:20:54.620 | so we can deliver for our customers, right?
00:20:57.140 | We have only 8,000 customers, only 8,000,
00:21:00.500 | and if you think about 8,000 customers,
00:21:02.820 | 10 billion ARR, you can do the math pretty fast,
00:21:05.580 | but we have only 8,000 customers,
00:21:07.540 | and so I'm obsessed, like OCD level obsessed,
00:21:11.420 | that if you come in and say,
00:21:13.380 | here is what it can do for your use cases,
00:21:16.120 | I will listen to your pitch every day, okay?
00:21:18.780 | So that's one, and I have enough money to spend,
00:21:22.820 | if we can serve our customers better.
00:21:25.540 | And number two is, we have a great go-to-market team,
00:21:29.100 | so besides our engineering, AI, science, research team,
00:21:33.020 | we have a great go-to-market team,
00:21:34.960 | and if you have something coming back to the buyer,
00:21:39.960 | next door down person, or two doors down person,
00:21:43.460 | and you want to leverage, CIO is our prime buyer.
00:21:46.740 | If you think about CIOs, 10, 15, 20 years ago,
00:21:49.660 | till ServiceNow came, CIOs were serving other C-suite.
00:21:54.660 | Hey, for sales, I need to put Salesforce in,
00:21:58.000 | for marketing, I may need to put Adobe in,
00:21:59.940 | for finance, SAP, I need to put for the CFO.
00:22:03.500 | We were the first platform,
00:22:05.020 | we said this is the CIO's platform.
00:22:07.540 | So if you say that CIO is your buyer,
00:22:09.720 | and you want access to the CIO,
00:22:12.900 | there is no better company to partner with
00:22:14.540 | than ServiceNow, right?
00:22:16.020 | The two companies that really sell to CIO well
00:22:19.120 | is ServiceNow and Microsoft.
00:22:20.980 | These are the two companies that sell really well.
00:22:23.420 | - Brilliant.
00:22:26.260 | I'm sure quite a few people in this room
00:22:28.020 | are interested in that $1 billion of cloud spend.
00:22:31.260 | All right, we've got time for,
00:22:32.900 | I'd say, let's say three to four questions.
00:22:36.060 | Michelle, hello.
00:22:37.540 | - Thanks, Vijay.
00:22:39.180 | Many of us have dreams of an Act Two.
00:22:41.700 | Some of us might already be thinking about Act Two
00:22:44.420 | in terms of a product,
00:22:45.940 | any advice in the early days of how you think
00:22:49.060 | about resourcing and philosophy around experimentation
00:22:52.460 | versus intentional bets around Act Two's product development?
00:22:56.820 | - Correct.
00:22:57.660 | So I will share a story that the reason
00:23:01.980 | our IPO was very mid is because
00:23:05.820 | people said that our TAM was only 1.8 billion.
00:23:13.260 | So one of the industry analysts said,
00:23:16.300 | "These guys are not gonna do well,
00:23:18.660 | "and their TAM is limited."
00:23:20.780 | Same thing happened to many, many companies
00:23:23.220 | where people say, "I don't know if the TAM is there."
00:23:26.140 | So for us, that actually created a chip on our shoulder
00:23:30.140 | because we believed that the TAM for ServiceNow
00:23:32.340 | was a lot bigger than what the industry analyst community
00:23:36.180 | said it was, which was sub-2 billion, okay?
00:23:38.820 | And this was in 2012, not too long ago.
00:23:41.460 | So one thing is on your core, core being core,
00:23:45.780 | you really, really have to understand what is the TAM,
00:23:48.780 | which is an art combined with science,
00:23:51.580 | before you start saying, "I wanna go multi-product.
00:23:55.000 | "I wanna go now to different buyers
00:23:56.820 | "or the same buyer, but multi-products,"
00:23:58.740 | whatever the strategy you wanna go after
00:24:00.940 | with the buyer access, but the core has to be core.
00:24:04.220 | The reason you exist is for something.
00:24:06.540 | You try to solve a problem.
00:24:08.460 | So really understand the TAM behind that core
00:24:11.540 | and then figure out, before you go into the next act,
00:24:15.580 | why are you really going after that next act?
00:24:17.780 | So we prevented going after the next act
00:24:22.780 | till we hit one billion in ARR.
00:24:25.940 | And then overnight, we flipped it,
00:24:27.740 | and we said, "We are gonna go after these three
00:24:29.220 | "buying centers, security, HR, and customer service,
00:24:31.960 | "in addition to IT, and here is the go-to market for it,
00:24:35.640 | "here is the buyer for it,
00:24:37.140 | "and we are gonna rely on CIO
00:24:38.500 | "to make introduction to those buyers,"
00:24:40.220 | because we nailed the CIO.
00:24:42.120 | So core has to be core,
00:24:44.220 | and you really have to understand the TAM
00:24:46.260 | before you say, "I'm now taking,"
00:24:48.500 | because it's so easy to say,
00:24:50.640 | "Oh, typical thing I hear from entrepreneurs,
00:24:55.020 | "CEOs, smart people like yourself,
00:24:57.160 | "hey, we have a great product,
00:25:00.220 | "but I don't have a great sales team."
00:25:02.360 | And then they flip constantly chief revenue officer, right?
00:25:07.360 | They do, some of you probably do.
00:25:10.480 | And I always, when I'm asked for advice,
00:25:14.080 | which rarely happens, but when I'm asked for advice,
00:25:17.000 | I say, "What problem are you really trying to solve?
00:25:21.680 | "Is it the chief revenue officer,
00:25:23.460 | "or you really don't know what product you are building?"
00:25:27.060 | So that focus, that maniacal focus
00:25:29.880 | on main thing being the main thing,
00:25:31.720 | and what is the TAM really in there,
00:25:33.640 | before you pivot, or before you go to second act,
00:25:36.360 | is something that we look out for.
00:25:37.840 | So that's what we learned.
00:25:38.720 | We said main thing should be the main thing
00:25:40.800 | till it hits billion,
00:25:42.400 | before we go to the three other things.
00:25:44.680 | - Any Sequoia company can discuss
00:25:48.840 | act two after a billion of ARR, is what I'm hearing.
00:25:53.200 | And I think a very powerful insight there, CJ,
00:25:55.640 | when we get the next question set up,
00:25:57.560 | is just the power of the CIO.
00:25:59.960 | I think a lot of people overlook that,
00:26:01.800 | that HR and security would look to them.
00:26:04.600 | Charlie.
00:26:05.640 | - I have a question, which is just,
00:26:07.200 | ServiceNow is a broad platform with many capabilities.
00:26:11.000 | You have other ways you interface with customers,
00:26:13.360 | like customer support.
00:26:14.720 | How do you think about prioritizing
00:26:16.800 | where you want to integrate AI?
00:26:18.620 | - So, it has been hard in figuring out
00:26:24.520 | where AI could truly disrupt a use case, right?
00:26:28.320 | Because that's always the hardest thing,
00:26:30.240 | because it's still very bleeding edge.
00:26:32.200 | And on the buyer side, if you're talking to a large bank,
00:26:36.600 | if you're talking to a large government, right?
00:26:39.280 | And I'll share one story on this.
00:26:42.760 | So US public sector, which is federal, state and local, okay?
00:26:47.760 | You have to invest a lot in US public sector
00:26:51.440 | for certification of your product.
00:26:53.760 | The cloud, you know, Microsoft has regions
00:26:57.240 | for IL-5, IL-6.
00:26:59.400 | So our chief revenue officer came to me and said,
00:27:02.200 | "We want to go all in on US public sector."
00:27:05.440 | And I said, "Okay."
00:27:06.520 | And we had to invest 100 million plus in infra
00:27:11.520 | before we can start really making money
00:27:14.220 | in US public sector.
00:27:15.880 | And now, whether it's US Army, US Navy, Air Force,
00:27:19.760 | to all public sector institution, even on civilian side,
00:27:23.600 | are all ServiceNow wall-to-wall customers.
00:27:26.560 | And coming back to the question,
00:27:29.400 | we always try to figure out the pain point
00:27:33.200 | and can AI really disrupt that use case
00:27:36.880 | in a positive way that customers gets higher value
00:27:40.520 | from ServiceNow.
00:27:41.960 | So if I infuse AI in a use case,
00:27:44.040 | because not all use cases are created equal, right?
00:27:46.600 | You have a product for multiple use cases.
00:27:48.960 | Not all use cases are created equal,
00:27:51.160 | but which use case will really provide higher value?
00:27:55.080 | Because when customers spend money on you,
00:27:57.680 | all they are looking for is how much value
00:28:00.160 | I'm gonna get out of this investment.
00:28:02.040 | And that software ROIC, including AI ROIC, is hard.
00:28:06.800 | Like right now, we are trying to tell everybody that,
00:28:10.560 | okay, we have gen AI infused in ServiceNow products.
00:28:14.180 | Number one question is, how much will it cost?
00:28:17.200 | And what's the return I'm gonna get?
00:28:18.760 | And if you're not doing outcome-based selling,
00:28:21.200 | so if you're not doing outcome-based selling,
00:28:23.840 | it's like you are another guy coming in there
00:28:26.240 | and giving the AI pitch to the customer.
00:28:29.120 | And truth is, nobody gives a shit.
00:28:31.260 | I mean, they don't.
00:28:32.140 | Because you have to be very, very clear and specific
00:28:35.320 | on here is where you will get the ROIC on that.
00:28:39.140 | So wherever the highest ROIC is, that's where we prioritize.
00:28:42.140 | That customers can say, okay, I could see,
00:28:45.100 | if CJ is saying $10 million productivity
00:28:47.280 | for this large bank, most likely,
00:28:50.200 | because if CJ is gonna be $3 million,
00:28:52.160 | but $3 million is still better than zero.
00:28:54.440 | Yeah?
00:28:55.280 | - Andy.
00:28:57.800 | - You have such a close understanding of what the CIO wants.
00:29:02.120 | - Yeah.
00:29:02.960 | - Beyond HR support services, security, IT,
00:29:08.940 | what's the next set of use cases
00:29:11.560 | that the CIO is super excited about
00:29:13.480 | beyond what ServiceNow is currently harvesting?
00:29:17.200 | - Yeah, I would say, if you think,
00:29:20.040 | and one thing you all should know,
00:29:22.120 | this is what I learned, is in my seven years
00:29:25.520 | at ServiceNow, they constantly told me
00:29:29.440 | about CIO being chief irrelevant officer, okay?
00:29:34.440 | And they said, all the power is with developers.
00:29:38.040 | And CJ, you are selling to the wrong door.
00:29:41.200 | You need to be like other people and sell to developers.
00:29:45.240 | Yes, developers will buy X, Y, Z, all that is fine.
00:29:48.960 | But the irrelevancy of CIO has been exaggerated.
00:29:53.960 | And right now, CIO is the most technical person
00:29:58.160 | on a C-suite at most of the large companies
00:30:01.920 | which are your buyers, right?
00:30:04.240 | You may have a CTO, the product person, right?
00:30:06.640 | The product tech person, and you have the CIO.
00:30:09.320 | But product tech person is focused on innovation,
00:30:11.840 | not about what they will buy from you.
00:30:14.040 | So coming back to CIO's irrelevance,
00:30:16.120 | has been exaggerated and over-exaggerated year after year.
00:30:20.040 | And now, we mainly sell to Fortune 500.
00:30:23.880 | In Fortune 500, the CEOs are only asking CIOs,
00:30:28.120 | give us the AI roadmap, give us this,
00:30:30.080 | give us that, and so on.
00:30:31.040 | So that's number one, that CIO is still very relevant
00:30:34.000 | and very important, okay?
00:30:36.360 | It's not that developers are not,
00:30:38.120 | but if you're selling to developers,
00:30:40.160 | I mean, I was one, very fickle people, right?
00:30:43.200 | They churn and sell thing, and someday they like this,
00:30:46.920 | and I read this on frickin' Reddit, and now I like this.
00:30:49.560 | (audience laughing)
00:30:50.400 | That's a hard thing to sell.
00:30:53.000 | So CIOs right now, with the state of economy today,
00:30:58.000 | are focused on two main things.
00:31:02.540 | One, where can I take out the cost,
00:31:06.200 | just enterprise-wide using technology?
00:31:09.360 | So where can I take out the cost?
00:31:11.240 | And second, how can I help in the path to revenue?
00:31:16.240 | If I can help on the path to revenue,
00:31:18.360 | whether it's quote to cash,
00:31:19.840 | or whether it's any part of the sales,
00:31:23.320 | front end or front office, CIO is constant.
00:31:27.000 | Like today, I'll tell you at ServiceNow,
00:31:29.320 | it still takes us a time when we propose
00:31:32.160 | a bill of material to a customer,
00:31:34.440 | and customer says, well, I want to change quantity here,
00:31:37.120 | quantity there, then we have to re-spin the order form,
00:31:40.800 | make sure it's in SAP, make sure it's in our CPQ system.
00:31:44.100 | That process is still three, four hours sometimes,
00:31:46.200 | and at the end of quarter,
00:31:47.120 | three, four hours feels like eternity.
00:31:49.200 | So how can we do that fast?
00:31:52.800 | Well, CIOs right now are the two.
00:31:54.960 | The CFO and CRO are their biggest stakeholder,
00:31:59.380 | and if your product is in path to that,
00:32:02.100 | you'll always get that CIO meeting.
00:32:04.300 | But you have to make it really quick,
00:32:06.320 | because I talk to seven, eight CIOs a day,
00:32:09.960 | and that's what I constantly see on the pattern matching.
00:32:12.840 | - Excellent, we have time for one more question, Peter.
00:32:17.000 | - All right, you talked a lot about use cases.
00:32:19.840 | I'm curious if you have any stories of use cases
00:32:22.080 | with AI that worked really well,
00:32:23.780 | and use cases with AI that really did not work at all.
00:32:26.720 | - Yeah, so one of the things I would say,
00:32:30.080 | trying to read documents and understand documents,
00:32:37.120 | which is accounts payable, invoices, and this and that,
00:32:41.000 | it is still very hard for AI to crack that,
00:32:43.800 | and we have tried multiple different ways,
00:32:47.280 | and people talk about OCR and this and that,
00:32:50.300 | and that is a junkyard of technologies that we have tried
00:32:55.160 | and not been able to crack through,
00:32:56.860 | because if you can automate the paperwork,
00:32:59.800 | invoice matching, and other things,
00:33:02.000 | there is still a lot of dollars to be had on productivity,
00:33:04.560 | on booking the revenue, and so on.
00:33:06.760 | Where we have seen the most is simple things
00:33:10.400 | like predicting X for our use cases,
00:33:15.280 | specifically for our use cases.
00:33:16.760 | So for example, say a large bank.
00:33:20.100 | Their policy is that if you use your computer,
00:33:23.360 | every three years you can refresh your computer,
00:33:25.320 | say your Mac.
00:33:26.140 | We can do that now with AI, and we can say,
00:33:29.160 | "Hey, Julie, you're four months away
00:33:32.680 | "from your entitled computer refresh,
00:33:36.040 | "and we have already notified IT,
00:33:38.220 | "and you will get your new computer on that third,"
00:33:40.480 | so the depreciation and all those schedules work,
00:33:43.240 | and say yes if you agree,
00:33:45.560 | because we still want human in the loop,
00:33:47.120 | because nobody likes.
00:33:49.080 | So far, companies don't like if we just make the decision.
00:33:51.980 | That's where we see that when we see the pattern matching
00:33:55.160 | and can we predict better and make it easy
00:33:58.040 | is where we are seeing the highest leverage
00:33:59.940 | on even generative AI.
00:34:01.360 | - Amazing.
00:34:03.720 | CJ, this is fabulous.
00:34:05.400 | I wanna tie together some findings here.
00:34:07.880 | Really, this is unique in that we had two separate areas.
00:34:11.320 | First is application layer.
00:34:12.920 | We've been talking about this throughout the day,
00:34:14.760 | and here's someone who has done it
00:34:16.160 | incredibly successfully at massive scale.
00:34:19.360 | You talked about who's the buyer,
00:34:21.880 | what is the size of the prize,
00:34:23.920 | and who are you competing against,
00:34:25.800 | and also the overlooked customer.
00:34:27.720 | In this case, the chief irrelevant officer.
00:34:30.720 | You said that really actually is incredibly powerful.
00:34:33.360 | I'll tell a quick story, which is at dinner with you
00:34:36.440 | and Bill from ServiceNow a little while ago,
00:34:38.880 | you guys said to me, hey, you get to a building,
00:34:41.360 | you go to the elevator, what floor do you go to?
00:34:44.480 | I said, I don't know, what floor?
00:34:46.100 | The top.
00:34:47.440 | Okay, great, you get off at the top floor.
00:34:49.700 | Where do you go from there?
00:34:51.040 | Bathroom?
00:34:53.240 | No, the corner.
00:34:55.220 | That's the person who's the buyer,
00:34:56.640 | and that seems like it permeates your mentality.
00:34:59.400 | And two, actually focusing on the customer.
00:35:03.380 | Each of you in this room, if you succeed,
00:35:05.920 | will try to sell to the likes of ServiceNow,
00:35:08.720 | a billion dollar cloud spend business.
00:35:11.480 | And you gave us a great guide here with AI.
00:35:13.480 | For you, you're looking at small language models,
00:35:15.720 | not just the biggest ones.
00:35:17.040 | You're looking at cost.
00:35:18.240 | You're looking at open source.
00:35:19.720 | 1% gross margin matters to you.
00:35:21.360 | That's how you build an 82% gross margin,
00:35:23.800 | 160 billion dollar business,
00:35:25.540 | and actually understand the customer.
00:35:27.880 | Don't just go to the marketing website
00:35:30.000 | and try to guess what ServiceNow does.
00:35:32.560 | Talk to people who've built it.
00:35:34.040 | This is fabulous.
00:35:35.160 | CJ, any parting words for the team?
00:35:37.160 | - Nothing, it was a pleasure,
00:35:38.600 | and I'm always around if you want any words of wisdom.
00:35:42.160 | I've made a lot of mistakes as well, too many,
00:35:44.780 | in terms of scaling ServiceNow.
00:35:48.320 | From a product perspective, engineering perspective,
00:35:50.680 | which use cases you prioritize, which you don't,
00:35:53.120 | how do you ring fence the team
00:35:54.620 | when you go for the second act,
00:35:56.560 | and really, really get focused?
00:35:58.920 | So, lots of mistakes as well,
00:36:01.160 | but we are constantly learning.
00:36:02.560 | - Fabulous.
00:36:03.400 | CJ, one last question for you.
00:36:04.240 | Will you do stand-up for us later on tonight?
00:36:06.560 | - No, I will not do stand-up today.
00:36:08.440 | It's my stand-up when I used to do it.
00:36:12.040 | Nowadays, I'll get canceled very fast,
00:36:13.880 | and Bill will not appreciate that, so I will not do it.
00:36:17.080 | - Well, there we go.
00:36:17.920 | That one's off record.
00:36:18.740 | Thank you, CJ.
00:36:19.580 | - Thank you.
00:36:20.420 | (audience applauds)
00:36:23.420 | (upbeat music)
00:36:26.000 | (upbeat music)
00:36:28.580 | [BLANK_AUDIO]