Salesforce, Inc. (CRM) Earnings Call Transcript & Summary

Marc, this is -- we're going to make it memorable. This is our last fireside chat, I'm told.

Well, I'm just trying to get over this news that you just told me and it's upsetting.

Well, it's been a good run.

I mean, you're retiring?

On the last day of your fiscal year. On the last day of your fiscal year. I timed it. I timed it. Yes.

I mean, it's unbelievable. How long have you been now at Goldman Sachs?

5 years.

And before that?

Total of 31 years.

31 years. Congratulations.

Thank you. Thank you very much.

Amazing career.

Thank you. Thank you. Didn't mean to elicit claps, but...

You said, you just want more time in your life to do things?

Open page. It's a blank page. We don't know yet.

Like getting a haircut or something?

Yes. Not time for that. But I want to, first of all, thank you for what you've done for the industry. You created the SaaS industry. You created the cloud. And I still cannot get over the fact that when I first met you was the first Dreamforce 2003, and you were doing $50 million in revenue, and you had this audacious dream to build what we call now a SaaS company, a cloud company. You had the goal that you wanted to be as ubiquitous as windows and as sticky as SAP. Here we are, I mean, 22, 23 years later.

What is ahead for Marc Benioff. And what is ahead for Salesforce? What does the company look like? This is the rate of -- the time of tumultuous change and all kinds of questions, what's ahead for Marc?

Well, I got a haircut. So I'm okay. I -- well, first of all, congratulations, Kash. You've done an incredible service to our industry and our community and to the company, and we're also grateful for everything that you've done. And how many earnings calls?

125.

Yes. It's amazing. I -- I mean, if you want to take it at the tippy top, I guess, when I get up every day, I've never been more excited about my job. I really think that right now, we are at the kind of beginning of something that is going to be the biggest transformation in enterprise software that any of us have ever experienced. And already, at a high level, I'm excited about where Salesforce is, what we've become, where we are. I think at that point, when we met, we were doing something like $50 million in sales or something crazy like that. And I mean you can do the math. You can look at the numbers from this quarter and see where we are and probably only spending a couple of years in the $40 billion and on our way into the $50 billion. And I don't think like when we had that conversation that we kind of saw that kind of clear growth trajectory. So that was kind of a moment then, but now is a moment also. And I think a year ago, if I look back when I was here, talking with David, we were just at the beginning, starting to talk about this kind of revelation that we had where we kind of really saw that so many customers were doing so many things with artificial intelligence and so excited, but they were still kind of grappling with the value proposition, what are they doing, what is the outcome going to be? And we saw that even manifest into this MIT study that appeared in the last month or 2 where so many customers have spent actually a lot of dough, but haven't completely got the outcome. And a year ago, we were kind of on our track. We had just really -- we were kind of coming into the Dreamforce zone, if you remember the conversation. We knew what we wanted to create, but it was still months away from actually even getting the first initial round of code into the market. So now we're about 9 months after delivering the first version of Agentforce, which is in service. Now, at that time, I wouldn't look back and go, "Oh, this is that kind of clear revolution of what's going to happen." But now I feel like I can take a few minutes and kind of give you like where I think things are going to go over the next few years. And I think that for us, we've become our own best example. And that's different because I think if we're not doing this first and really showing what's possible, we're not going to be able to really motivate all of our customers to be able to achieve it. So -- and it's not that we don't have -- now have 12,000 customers or something like that already starting to implement Agentforce. It's this idea that this is very different. So a year ago, we had, I don't know, 6,000 or 7,000 or 8,000 support agents, whatever it is. And we had our support application, and we're managing our information. And there was Einstein in there, as you know. We have all different levels of AI going on. But this idea that we could somehow harness not just generative AI, but kind of a different way to kind of manifest the technology. This was an idea. Today, it's not an idea. Today, at Salesforce, in the last 9 months, there have been 1.5 million conversations done by our support organization by what's now about 4,000 humans who are doing customer support. But 1.5 million conversations have been handled by digital AI agents. And there's a orchestrator that is orchestrating between the humans and the AI, keeping them all in sync because the AI cannot do everything, but it can do some things. And the humans and the AI working together can achieve an incredible outcome. So what's cool is that the CSAT score of the AI agents and of the humans is about the same. And that is also a huge surprise to me. So when we look at kind of what we've traditionally called our Sales Cloud, now in our mind, we say, no, this is actually agentic -- sorry, what was our Service Cloud. I'm going to get to our Sales Cloud in a second, is now agentic service. And that idea that we have agentic service where humans and AI are working together to deliver customer service. At Salesforce, that has been transformational. And we had an all-hands call yesterday with our employees of -- 75,000 to 80,000 employees. And really explaining, number one, yes, we're building these products. But two, we're also reshaping our company to be an agentic enterprise. So we're showing what we can build with the tools and we are going to be #1, our own best practice. In the second example, which is another huge surprise and I mean I could keep going for a while is sales. And in the last 26 years, maybe there have been between 20 million and 100 million people have contacted Salesforce, who we did not call back. We just didn't call them back. And we didn't have the people to do it. But now we have an agentic sales and that is linked very tightly with our sales organization. We have about 15,000 salespeople and then we have this agentic sales as well. So the extension is incredible. So this kind of Sales Cloud has evolved into agentic sales. And then each one of these products that we look at, whether it's sales or service or field service or Tableau or Slack, in each and every case, you can see that it's not just about the traditional application, working with humans, but then humans are working with agents as well. And that is a big change for us. And it's not something I think that was going to happen so fast. Even like in my home, I have a Airstream trailer outside and that Airstream trailer is kind of hooked into my power supply in my home on a device made by one of our customers called Eaton. And Eaton has a big field service organization, they come out and they repaired and so forth. And they've used our field service product, and you could see it and go in the App Store and get it. But now when that field service agent comes out, there's an agent as part of the app. Not only is it kind of here's Marc Benioff, here's his home. Here's the device, here's what it's connected to. This is the whole service history, but also the agent is able to work with him and say, "Yes, here's how to improve it, here's how to make it better," all of those things. In all of these cases, whether it's sales or service or field service, and I can keep going, it's humans and agents working together to create that customer success. And that is what is really exciting. And when we get to Dreamforce on October 14 and I hope that you'll come.

I will be there, of course.

You'll see, I think, just about every single one of our products.

I've been to every single Dreamforce since the first Dreamforce.

I am very grateful for that, by the way. Yes. And Trevor, would you bring in my phone for just 1 second. And I'm just going to show you because one really amazing thing is, if you just look at my phone, I'll just show you like I run my business on Slack. You know that. It's great. We have 1 million customers on Slack. We have 150,000 or something companies on Salesforce. But if I just come on here and I just want to kind of look at what's going on, on Slack, okay? And then you can see right here, if I just go to Slack and then I just go to Agentforce. Here, you can just see -- you can take it. You'll see like there's dozens of agent forecast. You're not going to see anything. Don't worry. I wouldn't be handling your phone -- I wouldn't hand it to you. But what you can do though is you could renew some customers, you could sell something, you could even get into the HR benefits, and you can see how the agents are just running right inside Slack. That idea, again, now that I'm now working with the agents myself so that you can see I have a CEO agent, I have a sales agent. If I go here, I have all my agents, so I can get in, I can even get in here. I can operate every aspect of the business right from here. That idea kind of where we used to talk about, I can run my business from my phone. But now I'm kind of running it in partnership with these agents. So on one part of my business over here, I've got service, and I've got, yes, my 4,000, 5,000 service agents, but they're working with my thousands of service agents as well. I've got my sales force, but they're working in partnership with the sales agents. And I'm working with these agents. And in every aspect of my business, this is now happening. And that's powerful. And it's making us more profitable, more successful, giving us higher customer satisfaction. And now I think we have clarity that for each one of our customers, and I just got off the road, I was on the road, as you know, for 8 weeks in Europe, met with hundreds of customers. Each one has the opportunity to go through the same kind of organizational transformation that I'm going through to be more profitable, to have better cash flow, to have higher customer success, to use technology in this way that's going to automate me in this incredible new way. It's going to require a lot of change management for those companies but kind of very early examples of the success of these customers is amazing. So that vision of an agentic enterprise, a vision of humans and agents working together across every line of business, this is kind of my belief what will be true for our whole industry going forward.

And the future of the software industry, has been called into question because of AI. Is someone that -- I mean you started Salesforce in 1998 or not, was it '98? Yes. And you worked at Oracle, you went through the whole client server transition. You actually -- when I met you, you explained to me what the web browser is going to do to the enterprise software industry, and you are the first executive that I knew that was able to put out web browser front end on top of a Drab user interface and make it look so pretty, right? When I look at AI, I feel like AI is a new UI. But then people tell me that, well, yes, it's going to take over software. You don't need applications anymore. I can custom build an AI that can do forecasting and planning and this and that. What do you think of that?

Well, I think that we -- you can see how for customers, they've gotten very entranced by these ideas should they have their own models? Should they have this? Should they build this themselves, what we call it, should they DIY their AI. And one of our customers is a bank, not Goldman Sachs, but a large bank, one of the largest. And I was talking to the CEO and they're like, "Listen, we have the PhDs. We have this, we're doing this. We're rolling our own. We're going to make this work." . And I'm like, okay, but in your wealth management, we're already working with them with Agentforce and take a look at these results. And then when it actually got to that level, they could say, in this case, this ability to use a platform that's going to give you 3 key things: First, it's going to give you the applications that your humans need because at your bank, at my company, at every company reflected in this room, there's a lot of employees who are going to be automated and those automations can happen through various levels of applications. Maybe some of the user interfaces on those applications are changing. But you can see even in the applications like I just posted on my X-feed this weekend, Slack, for example, this is still a very powerful application that I'm using every single day. I still need to direct message my employees. I still need to collaborate. I still need to look at all my analytics. I still need to understand and get access to all my customer records. As you said, my forecast, but I'm working with agents, I'm working with the AI to achieve my success. I've been doing this myself for quite a few years. But to get it working actually part and parcel as part of the application. That is what's really been so exciting to me. And that, I think people got a little bit confused thinking like, "Oh, wow, this is so exciting. Does that mean we're going to just disconnect from the rest of the organization." Well, then how are you going to manage all your sales folks or all your service professionals or yourself? And you have to have a platform where humans and agents are going to work together.

How do you see the -- you also not only put the -- the first person to put web browser on top of enterprise software, but you also had the SaaS model, Software as a Service. Remember that one? So now people tell me that SaaS is kind of a flat industry. It's all about consumption. How do you monetize? What is the business model of a Salesforce look like 1 year, 2 year, 3 years down the line? What do you see just consumption versus seats play, interplay in your base?

Well, I think that seats have a role like, for example, even like I think you had Sarah speak earlier on ChatGPT like you can see that's a seat-based model, right? Like you probably have like subscribed for $20 a month or $200 a month. So...

The $20, not the $200.

The seat-based model is a model that will continue in software. Consumption is also model. Like, for example, we have our data cloud, that's a consumption-based product. We have our Commerce Cloud, that's a consumption-based product. Our e-mail marketing, we do 11 trillion e-mails a year that we sent out, that's trillion with a T. That is all -- currently, for example, that's all one-way conversations, right? We're -- email -- sending out these 11 trillion e-mails, but before the end of the year, and you'll see this also Dreamforce, that will transform that at the end each one of those e-mails is going to be an agent, right, that you're going to be in conversation with. So that -- as a consumption model, that will continue. And so you'll have a combination of usage models, consumption models, transaction models, all of those things. And for a large company like Salesforce, we're not a small company and we're not a single product company, we're going to have many different models that are going to work together. For our customers, especially our large customers, what they want now is they want -- when I'm working with them, they want an agentic enterprise license agreement. They want us to be able to come in and kind of give them one price with everything bundled together over 3 years. And as we're starting to deliver those or as they're getting ready to go through their transformation, that's what's exciting to them. And those are obviously extremely large agreements. And you saw that those extremely large agreements grew very dramatically in the quarter, and I expect for this year, we're going to see a lot of that. And the customers that I'm meeting with directly in almost every single case, that CEO, they have a fever for transformation, but they don't have a trajectory. They haven't really had -- they know they could get more productivity from AI, they know they can have better KPIs. They know they can be more profitable and in all these things. But in many cases, they don't know how to get from A to B. And our job is to be their trusted adviser and say, "Here's where you are and here's where we can get you." And we have examples of customers in your industry that we can show you, but let's start with customer 0, us. And we'll look at our numbers and look at what we've been able to do on productivity and all these things and how -- this is how we're going to get you there as well. So that is kind of the last model, which is they're still going to want some kind of ability to kind of receive it all in one agreement.

Got it. So you've got now 12,000 Agentforce customers. If this is to scale, how do you see the underlying technology changing to ensure that this foundational platform can support 15x, 20x. We're talking about -- Thomas Kurian was earlier here today. And practically, every 1 of the AI native is talking about 10x, 20x. So to support that kind of growth in transactions how do you scale this agent technology going forward? Just as you scale the SaaS platform from you wanted to be 1 million users, you ended up being like 50 million, 60 million, 70 million. And we went through a learning experience, right?

That's -- I would say there is one part of this where we're better lucky than smart. We were already going through before really the LLM revolution kind of a data transformation of our products. We had decided that we would do -- first of all, we would take all of our core applications. And we would start to -- especially the ones that we bought, acquired over time, that we were ready -- we were going to integrate and we were going to create one application platform, one application layer. So we're going to rewrite Tableau. We were going to rewrite MuleSoft. We are going to rewrite Commerce Cloud. We are going to do all this and bring it together so that it could have fluidity at the application layer. We really had a vision that we could deliver that and deliver it at a level of speed and scale that would offer our customers a level of functionality like they've never had before. But then we wanted to add one more thing to that. We wanted to add a data layer. We were early investors in Snowflake. You know that. I think, we made...

[ Sridhar ] was here, literally, yesterday, here.

I mean, it was a great investment for us, I think. Our return was something like $1.5 billion or more, something like that. One of our best from our venture portfolio, Databricks will be another one as well. We're very...

Ali is going to be here tomorrow.

Great. We're very inspired by those data companies and as we looked at them, we're like, it's even more powerful if we take our applications and we integrate them with the data cloud. So we love those companies. We partner with those companies, but we also want to have our own data cloud as well. We want our data cloud to federate easily to theirs. . So that means if you have a Snowflake or a Databricks or like you mentioned like a Google BigQuery or an Amazon Redshift or even now IBM mainframes and others, you can just -- even Workday, you can just hit the button. And our data cloud, it's as if the data is running in our system but we can kind of shadow it and keep it running in their system, if that's necessary. And it's an activation of the data onto our platform. So now our applications in that whole application layer that I described with all those kind of applications, that is now able to read all that data. No one else has done that work. And then it was a year ago that we said, "Well, maybe there's a third layer, an agentic layer at the top of that." So if you had an application layer, a data layer and an agentic layer, what would happen? But we are writing all of that for scale, for multi-language, multicurrency first. Maybe we're one of the only software companies that is doing that. A significant amount of our customers are small businesses and medium businesses, small businesses are like 0 to 200 employees, medium businesses from like 200 employees to a couple of thousand employees. General -- we call it general business from like 2,000 employees to 5,000 employees or very, very large companies, the Fortune 100 companies, we run so many of them. And the government and then ISVs. And those are like our 6 segments. So those 3 layers then need to scale across all 6 of those segments as well. But that idea that those -- and people ask me like, for example, in our example with Disney, we've been able to achieve in some incredible power with our agentic layer with Disney. One of the reasons why because for a human doing Disney's customer work, it's very complicated. They've got such a great product line that they have so many options. I'm going to go to the park. I'm going to make a reservation at these restaurants, maybe I have these allergies. I'm going to make a hotel reservation. I'm going to get a special promotion. By the time you get it all put together, you kind of need AI to kind of help you configure this product that Disney can offer you. But our AI has provided this kind of 93% accuracy. And the reason it gets that high is because it has context. And the context comes out of the data. If you don't have the data, you're not going to achieve that level of accuracy with the AI, and that has been a tremendous shift for our customers that we're able to apply the applications, the data and the agentic layer altogether. And I think that's -- I'll give you an example, just right down the street, Williams-Sonoma, based here in San Francisco. We did our first agent maybe 6 months ago. It's a service agent, it's a sales agent. It's working with their customers. It's been so successful they have now deployed it for every one of their brands. We want to be able to turn Williams-Sonoma into a complete agentic enterprise. And all of the same examples for me, I want to be able to take them and every single one of our customers and say, "Here, now all your humans and agents are working together."

Marc, if we are to look at -- for investors that may have not completely bought into this, what are the indicators that they should be looking for that you will be sharing with us to help gauge that, okay, you know what, I'm confident that Salesforce is a real player in the agentic world, in a SaaS plus AI world or seats plus consumption world.

Well, one of the things that's been very important to me is that we've gone through in the last call it, post-pandemic is this complete financial transformation of the company. So I think you know we're going to deliver about $15 billion of cash flow. That's been very exciting for us, very high levels of profitability, kind of now very squarely in the mid-30s. We're going to -- we have these kind of core financial metrics. But then we have a very strong traction on getting customers signed up on the agentic layer. Now the reason why that is, this is a logical extension of our relationship with these customers. So we have to take our 1 million Slack customers, 150,000 core Salesforce customers and get that agentic layer running with -- for all of them. And I think that they're very motivated to do that. But if we went back a year ago, I keep going back to the year ago example, we hadn't used the word agent before. We hadn't used the word agentic. I remember even when I was on Mad Money and I used the word agentic for the first time, which was I think it was about 9 months ago, and Jim Cramer said, "Wait a minute, what is the word agentic, what does that mean?" And that's where we were in the industry. It's going very fast. But that idea, yes, we should just get every single one of our customers signed up for that. So yes, there's going to be a certain amount of consumption out of our data cloud. It's going to be a certain number of transactions with our agents. It's going to be usage. It's going to be also growing a lot of our seats as well. We're still growing so many of our seats in all of our core clouds.

How does it happen? I thought AI was going to take away jobs in marketing. You're not going to have marketing jobs, not going to have customers support jobs, developer jobs. What is your view as someone that's been through these cycles where tech automates, introduces productivity and there's this doomsday scenario, it's going to take away jobs, and it doesn't happen, could it be real this time?

I'm trying to be that best practice myself, as I mentioned. We have 75,000, 80,000 employees. They're not all in the same seats that they were 9 months ago. I have already radically reshaped the company. So I have thousands of more salespeople, but I don't necessarily have thousands of more headcount. I've moved people around. And when I'm on that -- I said at all-hands call yesterday, my message to my employees also is it's time for all of us to kind of get to another level in our capabilities in AI in this core technology because where all of our roles are going to slightly shift including mine. But some of the predictions that have been made in these areas, I think, might be a little bit aggressive or may be made by people who don't run the companies because I don't know exactly what they're talking about. I don't -- I know that some jobs will change, but not all jobs are going to change.

Wanted to also ask you, when you talked about the financial discipline, is it -- I would imagine that there is a desire to reaccelerate top line growth rate because I mean, there's no better way to show improving trajectory than to show better top line growth rate. Is that possible for Salesforce?

Well, I think you've seen it, right, in the last couple of quarters that there is some acceleration happening, and I expect it to continue. I don't want to get too aggressive in my commentary, and I'm not giving guidance, because I just gave guidance, but I expect -- and I -- my goal is to move into -- back into the double-digit growth category even as we're kind of starting to enter into the 50s. Like I was saying that we -- as I said, we're all able to do the math and put these things together. But I believe that we are going to go through a huge investment change in our industry. I think that coming out of the pandemic, for a lot of software companies, including ours, a lot of software was sold in the pandemic. As an example, we've more than doubled the size of our company since the pandemic. You know that. So now we're getting close to tripling the size of our company since the pandemic. And when we went through that, there was definitely an overage of software that happened in the pandemic because there was such a -- you remember that it was like this kind of frenzy around it. And then there had to be an absorption, I think, into the customers. I think this was for most software companies, not just ours. And now we're coming out of it, but not just that we've kind of rightsized that. In fact, our management team just met a few hours ago to talk through this very point. But the next piece, okay, is -- and I think what we didn't see is now we're into that good flow with our customer, but now we want to take that customer and transform them. I want to be able to take every single one of those customers that I mentioned, whether it's a Williams-Sonoma or DIRECTV or Reddit or any of the customers that I talked about on the...

Or Brunello Cucinelli.

Brunello Cucinelli, every single one can benefit from this transformation. And this is going to be an incredible opportunity for each and every one of them to go through this. And so when I was with -- I mean, I'm not going to go through each and every customer name, but I was with a very large customer in Paris. And they run a very large industrial company. And it's somebody I've worked with for more than a decade, and they have a new CEO, and I'm working with that CEO and I just can walk in through the examples that I'm walking through you, but for their own business. As the new CEO, they have an incredible fire to do this transformation and they just -- it was crystal clear the benefit that it could bring them. I think in some cases, when we're selling technology, it's not as clear to the C level exactly what the benefit is. In this case, with an agent saying, we're going to help you become an agentic enterprise and here is exactly how we're going to do it and how fast we're going to do it, it's very motivating. And for me, as the CEO of a large company, it's been very motivating. And that's where I'm like, oh, no, the reason my pipelines are fuller and richer than ever before is because agents are filling them right now in ways that before I never have that ability. Instead, I was leaving highly qualified leads from my competitors on the floor. I just didn't have enough people to call back. So I was helping fuel the industry. Now I'm calling every single person back. I can have a conversation with every single prospect. I can service every single customer instantly. And the customer, if they don't want to be talking with an agent, they can immediately escalate to a human because it's seamlessly integrated between the application and the agent because it's the same piece of code. It's the same data set. It's the same agentic layer. That is really unusual. And I would say that a year ago, that was not as crystal clear to us as it is now, that not only are we just going to deliver a product but that we would transform ourselves and our customers while doing it.

Yes. And CEOs and founders like you that wake up every day thinking about this big transformation ahead and how you position the company for this new tech cycle, you guys are...

I will tell you, it was -- it's kind of a strange process because I get excited about building the technology and then what I also then like once I have the product, and it's kind of working, I like to get out there and work with the customers. And this summer, when I went to Europe and I lived in Geneva and then I was in Amsterdam. I'm in Paris, I'm in London. I'm in all these various places, working with these customers one-on-one, sitting just like we're sitting now.

Is that a change for you? Because you don't travel to see this many customers.

No, I do. It's something I've always done. It's what I really love doing. And I would say that our -- each and every customer, it's kind of -- they go to investor conferences just like this, and they get -- it doesn't matter what industry they're in. They could be in financial services. They could be industrial. They could be in tech. They could be in retail. They could be on any of these things. They get -- the first question they get, what is AI doing for you? What is your AI strategy? How are you going to create a better set of financial metrics for your business? How is generative AI going to change you? What does this technology mean for you? And for a lot of them, they've experimented that they've tried, that they've tinkered. They've done all these things, but they don't have the clear answer. And I feel like, no, this is actually where you are going to get the value right now. And that is what I think is very exciting. We always knew it was going to come aggressively in the sales and marketing area, but this is very clear step, set of actions they can all take. And I think that's going to be a huge accelerator on our business and our ability to grow our company. Look, you know because we've talked about it many times, I'm going to $100 billion in revenue. I'm not going to do it irresponsibly. I'm doing it responsibly.

Organic $100 billion. Organic?

Well, I think that organic is a very critical part of our business, and it always has been. But there's always going to be an inorganic part of our business as well, but it has to be a responsible part. It has to be done with the right level of discipline. Like I think Informatica is probably our best practice. We're buying it with the right level of metrics, the right ideas, but it's going to -- that Informatica acquisition, it really amplifies that AI foundation layer that is -- that AI foundation layer that we have, which is our data cloud, MuleSoft, Informatica and Tableau, that's fueling and embedded into all of those apps. Nobody has done anything like that for the enterprise. That is really awesome. So -- but when we bought Informatica, we weren't just willing to pay any price, whatever it had to be the right level, right capability. We're still looking at how do we bring it in correctly into the company but yes, inorganic has been a part of our business. Like Slack was a part of it. MuleSoft was a part of it. ExactTarget was a part of it. We had a lot of critical things. We've done more than 60 acquisitions. But now look at where we are from a financial metrics point of view, we should be able to bring those metrics forward as well. I think we still have some of the highest cash flow in the industry. I mean I don't know the numbers as well as you do, but I think $15 billion this year is pretty high in software. We're looking at exactly how do we get motivated to get back into double-digit growth and then how do we get to that -- our $100 billion goal.

On that note, I wish you really well, my friend. And thanks for the last 22, 23 years.

Thank you. I hope that we continue to work together, and I just want to -- I think for everyone in this room, I'll just tell you how grateful we are for everything that you've done for us for the last 30 years.

You're creating an industry.

Thank you very much.

Thank you so much. Thank you.