Episode Transcript
[00:00:01] Okay, guys, today I want to talk about Workday and we just announced a partnership with Workday, which I'll talk about in a couple of minutes. But let me talk about the company first because this was Workday Rising and last week was SAP Connect and then next week's is Unleashed. So we're going to talk about all this stuff.
[00:00:18] Workday is a very successful company. They're about almost a $10 billion company. They've been growing at 13 to 15% per year, very steadily over the years. Their revenue growth is almost a perfect cur.
[00:00:32] They have 11,500 large corporations as customers, 80 million or more end users using the platform. Very big in financial services and service related industries and other sort of high value industries. Software companies, a lot of educational institutions, government institutions, schools to some degree they're getting into media. And then, you know, some of the manufacturing industries use Workday, a lot of them use SAP.
[00:01:04] And the system was the first system in the world. And this is their pioneering beginnings that leveraged the cloud using an object oriented database that allowed every customer to be on the same version of Workday with a rules based workflow engine so that all of the data and applications and integrations that you built on top of Workday would be maintainable over time and you'd never have to upgrade anything because Workday would do that for you.
[00:01:36] And this has been a huge benefit for them. Their user experience in the early days was state of the art. Everybody loved was a totally integrated application. All the different parts of Workday work together, but it was a little hard to get data in and out. So over the years they've done a lot of things to create APIs, development tools.
[00:01:58] And then last year they acquired Sana, which is the company we've been working with for almost four years for Galileo.
[00:02:07] And now SANA is the new user experience to Workday. So all of a sudden Workday looks like an AI system. It looks a little bit like OpenAI when you want to use it, you kind of ask it a question and it does what you want and shows you what it's doing.
[00:02:23] And so what they launched this week was the formal introduction of SANA as the new interface to Workday.
[00:02:31] A series of seven major announcements, but 17 major multifunctional AI agents to use Workday.
[00:02:42] An opening of the platform using what's called the Workday data cloud, an integration and open platform for CLAUDE to access Workday. So if you're a CLAUDE customer, you can get data in and out of Workday very easily. A series of development tools to make it really easy for you to build your own agents in the Workday sauna environment and a relationship with us to partner on HR 2030 and to embed Galileo into the Workday environment also.
[00:03:13] And I will tell you a little bit about our relationship with Workday in a few minutes.
[00:03:18] Now, there's also a very interesting business story going on about Workday, which was written about Unfortune magazine and that is that Anil Bashuri, who was one of the founders of Workday, has stepped back, is the CEO and is very involved in the product strategy and the mission of Workday to be a safe and lawful environment for agents in large companies to act on your data without going rogue. So they've spent a lot of time and a lot of money on the Workday system of record. They just acquired an AI identity company that identifies agents so that we can provision them and manage them and monitor them, and an application called the Workday Agent System of Record. So you can manage all of these agents that you're building and buying into your company and keep track of them just like an organization model for people.
[00:04:15] Because agents, as we've been writing about, are a little bit like people.
[00:04:19] And the company's really fired up, I would say. And then the other thing that's big is they've obviously really enhanced their talent acquisition through their acquisition of Paradox, or really enhancing their learning through their acquisition of SANAA Learning, and announced a, I would call it sort of a fix, but an enhancement that when you apply for a job with a company that has Workday as their talent acquisition system, you create a profile and you can reuse the profile in any Workday company so you don't have to keep registering to use Workday to apply for a job, which has been a, a headache and a complaint all over the world for years about Workday's application for talent acquisition.
[00:05:01] So they really introduced a lot of stuff, talked a lot about the refounding of the company, the new management team and the AI strategy around it. And, you know, so there's really quite a bit of enthusiasm here amongst the Workday customers.
[00:05:18] Now. You know, Workday has been an exciting, fun, interesting, high culture place for a long time. The employees here are very loyal to the company. There's been a lot of turnover over the years, as with every company, but I think this is a very energizing moment for Workday. The customers here are very excited about the new agents. They're not sure how real they are. One of them, for example, is a workforce Intelligence agent that can take your workforce and your jobs and decompose them into tasks a little bit along the lines of what techwolf is going to help SAP do, so that you can delegate work to tasks, to agents and to people. An agent for the intelligent deployment or redeployment of people internally using planning tools, which is part of HR 2030, and a bunch of new agent tools for application development integration with outside tools like Claude and Microsoft Copilot. And it's a, you know, really a lot of stuff and a lot of momentum, a lot of new momentum from Workday.
[00:06:27] Now, what are we doing with Workday?
[00:06:29] We are really excited that we're doing two really big things. Number one, we're partnering with Workday on HR 2030. And we're going to be together using the Workday platform and our experience and relationships and their relationships with corporations to validate and test and give you tools, best practices, maturity models, examples and education on this new world of agentic systemic HR in the role of HR 2030. I've talked about HR 2030 in the past. HR 2030 is an integrated operating system of 161 agents that work together to turn HR into a growth tool and a growth operation for your company.
[00:07:23] And there's a fairly detailed white paper we wrote on it with Workday on their website. You can read a lot more about it from us. And we're going to work directly with workday in these 11,500 companies to help them understand this, bring them along the journey, help them prioritize their projects and sort out how they do this. We are not exclusive to Workday. We're still an independent analyst company. We still work with all the technologies in the market. But we can get a lot of research done with Workday and we're going to really use their help and they're going to use our help to tease out and explain and help educate people on this new world of Magentic hr.
[00:08:05] The other thing we're doing with Workday, which we're going to really launch publicly in a more formal way in November, is we're embedding Galileo into the Workday platform, all over the Workday platform.
[00:08:18] So anybody that uses Workday is going to an interface to talk to Workday, ask good questions, learn, turn Workday into a coach, turn Workday into an advisor. And all of the data and insights and intelligence that we've accumulated for the last 30 years is going to be available to every user of Workday, not just HR people.
[00:08:41] And those of you that Use Galileo, know what Galileo is capable of doing. And then we're going to add a lot more data into Galileo, which we'll talk more about another time.
[00:08:51] Now, competitively, Workday of course is in a very hot space.
[00:08:58] Companies like Hibob, SAP, Oracle, a lot of the mid market HCM companies are all building agentic. UKG are all building agentic interfaces into Workday. By the way, the other thing Workday is doing a lot of work on, we spend a lot of time with them is Frontline work.
[00:09:18] And they've obviously got Paradox and a lot of workforce management technology.
[00:09:24] And we're doing a lot of work with Workday and with Paradox on what we call the business of Frontline. And a lot of research we discussed at the meeting this week about how to improve the operational excellence and customer service of Frontline workers in all the frontline industries, which make up 60 to 70% of the United States workforce force, 80% worldwide. And that's another initiative that Workday is really investing in and we're partnering with Workday on that from again a research and advisory standpoint.
[00:09:57] So I mentioned in my last podcast or so what SAP is doing called the autonomous enterprise. You know, I think the interesting difference between the two companies strategies, even though they don't always compete because most companies, once they select one of these platforms, they don't, you know, bake them off against each other ever again, is that SAP has a vision of 80 or 90 independent agents in every part of SAP workdays building what I would call super agents that do more sophisticated things in more sophisticated ways.
[00:10:33] I won't get into a lot of that now. I'll be talking about it in some of the speeches we do. But Workday is trying to take a small lesson from Apple that sometimes it's better to do a few things well than to do a lot of things not very well.
[00:10:50] And I think that's a valid strategy because building an agent is not a small task.
[00:10:57] Every agent has to be extremely sophisticated in its own level of decision making, intelligence and context.
[00:11:05] And then it needs to interoperate with other agents and understand the context of the AI and non AI agents around it.
[00:11:14] So it's not as simple as one person sitting around and just building something and turning it on. It has to reflect the data and the processes or lack thereof of your company. I mean one of the companies I talked to at the conference this week said it's electric Utility said, you know, we don't have a lot of processes written down, but we have to Embrace AI. How do we discover and build the processes we need so that the AI can automate them? And so we talked a lot about this idea of process discovery and process redesign.
[00:11:49] And so that's going to be a big part of our AI future. And so these sophisticated agents that Workdays created are designed to help you model and use the processes you have. And because workdays had a workflow management engine for a long, long time, most workday customers do have a lot of fairly sophisticated processes. Whether they like them or not, they're there for reinvention.
[00:12:16] SAP's, you know, going down the same direction technologically, very different because SANA is a very different platform for what SAP is building.
[00:12:24] I think SANA in some ways is so sophisticated that SAP, I mean, Workday is not going to have to buy a company like techwolf to do skills inference, to do task inference, because SANA does all that.
[00:12:37] And the Sana Labs company is an AI lab, they're not just an application development team. So the sophistication in SANA is going to play out very, very well for Workday over time and it'll help them.
[00:12:54] But they're not the only ones doing this. Hi, Bob, for example, who's also embedded Galileo into their platform by the way, is probably encroaching on Workday at the low end. And then there's lots of mid market and new HCM vendors out there that have built new agentic HCM platforms. I won't mention all of them, but I'll be talking about them at the conference that, you know, people are still going to look at. But most of Workday's work is to enhance the value of the current customers and then add more value to those that are switching around and changing platforms and looking for new solutions in the mid sized market also.
[00:13:35] Okay, a couple of other things.
[00:13:39] The big issue that I think remains sort of open about all this AI stuff is the cost.
[00:13:49] And you know, the software vendors in the market want to make money at this and they've been essentially told by Wall street, if you're not generating revenue from AI, we're going to just hammer your stock.
[00:14:02] So whether this is valid or not, there's an intense demand to monetize the AI agents through consumption.
[00:14:11] Now the problem with that is nobody can predict the consumption cost of an agent because somebody might do something with it that's unpredictable.
[00:14:19] So we need either limiting tools, which workday has one, as does SAP, or we need business models and financial models to make this worthwhile if it costs 25 or $50 to book your travel through the Workday travel agent in Sana, Companies aren't going to use it and we don't know how much it costs. I don't even know if Workday knows how much it costs to run these tokens.
[00:14:49] And the way I look at it, Nvidia is effectively, and I don't mean this negatively, but they're effectively raping the market by charging so much for their chipsets that everybody up the stack has to pay a huge amount of money and charge us.
[00:15:07] Now, the consumer uses of AI, they've kept the cost down low because most consumers don't use the AI that much. So they can amortize the cost of a really high volume consumer against a whole bunch of consumers who don't use it at all.
[00:15:19] But in a corporate use where the CFO knows how much they're paying for the consumption and they see a small number of users burning up a huge amount of money, they're not going to like it. And I've had a lot of meetings here where this came up and I don't have an answer. And I think the vendors are going to have to think about how they bundle these credits or multi tiered prices for tokens into this.
[00:15:48] But what it really tells me, and I want to sort of remind you, is this stuff is not free.
[00:15:55] So even though it looks snazzy for a simple application, sometimes it might still be cheaper to use people.
[00:16:04] And we're going to have to figure this out over time in partnership with our vendor partners like Workday or SAP or somebody else. And the vendors themselves are going to have to become smarter about how much various transactions and agents actually do cost so that we can give clients advice and realistic planning on how much the AI is going to cost. And that gets back to me, to our HR 2030 strategy that, and this is beginning to come out in other research, but I've been singing this tune now for almost two years.
[00:16:41] AI is a tool for growth. It is not a tool for cost cutting. It may not cut costs at all. It might, but that's not really the goal. The goal is for you to learn how to do more and superpower your employees so that they can generate better value to their customers or patients or prospects or whatever they may be and deliver better products and services to your anointed customer base to allow you to grow faster or raise the price of your products because they're worth more and they're more valuable to the end users.
[00:17:19] You're not going to probably find the economics that compelling to get a 10% savings in how long it takes to book your PTO. I mean, workday made a big deal about the fact that the new PTO agent saves 80% of the time it takes to book PTO.
[00:17:35] And that's, you know, sometimes something like 1.6 million hours across all the people in workday who use Workday.
[00:17:42] I mean, that's really good. Nothing wrong with that. But what, how much does that little PTO agent cost? I don't know, maybe not very much, but it costs something, so that's not free.
[00:17:53] You know, maybe there's an easier way to do that. I'm not challenging any of these agents at all, but I'm saying that every customer I talk to is worried about this and they don't know how much this stuff's going to cost.
[00:18:04] So in the middle of this very incredibly exciting innovation that's going on, we have to figure out how we're going to pay for it.
[00:18:14] Okay, Second topic is the issue of Forward Deployed Engineering.
[00:18:23] Now, this also gets to my conversations with clients at Workday and at SAP.
[00:18:30] AI is not a set of tools or applications that you buy and turn on.
[00:18:38] It is like a probabilistic entity that learns from your data.
[00:18:46] So in other words, if you're a software company selling a product of some kind that's AI powered, whether you be Maki People or Eightfold or Paradox or Asana or anybody else, every one of your customers implementations is different.
[00:19:03] Every one of your customers is putting different data into it. Every one of your customers is getting different results out of that data. They might have clean or dirty data or heterogeneous or homogeneous data, and their ability and knowledge and expertise on how to tune and train the AI is going to impact the results they're getting and their satisfaction with you as a vendor.
[00:19:30] So forward Deployed engineering, or what we used to call professional services, is a critical part of this.
[00:19:39] And I heard this too.
[00:19:41] If I go out and I buy Paradox, which is a very focused application designed to do a fairly defined set of things, but a lot of these other applications are pretty broad and I turn it on and try to get it to work, and for some reason the data's not correct or the processing isn't giving me the right answers because the data is incorrect and I don't know why the data is incorrect or what there is about the data that's incorrect, I'm going to blame the vendor.
[00:20:10] So the vendor is going to have to. And I'm not going to hire Accenture because maybe Accenture is just as frustrated with or Deloitte with this thing as anybody else. They're not experts at this either.
[00:20:22] So I actually think another issue that's coming up in these big HCM systems, particularly for bigger companies, is the vendors, the providers of the applications are going to have to bring with them four deployed engineers who know what's going on to help you build the data infrastructure, the training, the guardrails and the surrounding integrations so that your agents behave the way they do in the demos. That's the same thing for SAP, that's the same thing for Oracle, that's the same thing for everybody.
[00:21:00] Now Wall street hates that they don't like software companies with professional services revenue because it's lower margin.
[00:21:08] But I would argue that maybe that's the way these companies costs justify.
[00:21:14] The token expense that we're going to pay is that we're buying a living, breathing, intelligent thing that was developed by the software company, but they're also there to help you learn how to use it. Now we're also doing this, of course we're not a super big company, so we can't help you and every one of your agents day by day, but we can teach you how to do this because we've learned a lot about it with our experience with Galileo. But.
[00:21:38] But I think that's another issue that comes up in a lot of these meetings is as mature or immature as all these AIs are, every company's implementation is different.
[00:21:48] So anyway, our partnership is very exciting. You can read all about it on our website and on the Workday website.
[00:21:56] I'm going to talk more about this and give you some other insights next week at Unleash in Paris.
[00:22:03] I think Workday is going to be in a very interesting position over the next couple years with the investments they're making. And I look forward to meeting a lot of you at Paris. If I didn't meet you here in Vegas over the last couple weeks. And lots more things that we'll talk about. We have the new version of Galileo is now out the Jupyter release. We're going to be doing HR 2030 workshops at Unleash. Any of you that would like help on your HR agent plan or education in general, join the Global HR Excellence Certification Program which is up and running and going like gangbusters. Or just call us up and we'll come on site and help you with your own strategy. Have a great weekend. Bye for now.
