Episode Transcript
Speaker 1 00:00:09 Hello everyone. Today I'm gonna talk about Workday and some of the big AI announcements this week and what Workday's response and strategy is in this area. So, as most of you know, there were a lot of big announcements this week in ai. OpenAI announced G P T four, which is the advanced model that has been powering Bing Chat. There's a technical paper on it, and if you read through it, you can see it is significantly smarter and more accurate than G P T 3.5. And I published an article a week ago from an M I T study that proves through several thousand people that users of G P T three, not four are significantly more productive for office related work than those who don't have access to it. So this open AI stuff is really, really good. Now, if you read the New York Times, you would think that it was an evil machine, but it's not.
Speaker 1 00:01:05 It is a very powerful neural net and it can analyze vast amounts of tokenized, textual data and images and predict and communicate and listen and understand your queries and talk to you. And you know, all this, I won't go into that. I'll talk more later about generative AI in general, but the big announcement this week was Microsoft. Microsoft announced the co-pilots for Microsoft 365, and I've been using Microsoft Tools since the 1980s. Believe it or not, I was working at IBM when the first PC was launched. So I used Viscal and Word Perfect and Lotus 1 23 and all those original tools. And when that stuff came out, it was just as disruptive as the co-pilot. In fact, I distinctly remember when I first started using those spreadsheets, I thought, wow, you know, everybody who does financial analysis and accounting is gonna go out of business.
Speaker 1 00:01:57 But of course they didn't. They just use the tools. So the co-pilots from Microsoft, which do amazing things, and you can read the article on that, are not going to eliminate any jobs. They're going to change all of your jobs and you're all gonna have to learn how to use it because if you don't, you'll be like somebody who doesn't know how to use Excel. So that was a huge announcement. And then of course, Google who somehow seems to not be able to catch up with open ai, and I'm not sure why, because I think they have all this stuff. There's something going on there pre-announced similar features in Google Workspace, but very weekly. So I, I think they're just struggling to figure out how to get their stuff out the door. I'll come, I know Google has a lot of amazing ai. I use Google Photos, which is just astoundingly good at organizing and finding content and, and that's the same technology.
Speaker 1 00:02:49 So that was a big series of announcements. Now, the other thing that I want to talk more about is Workday. Workday has an innovation summit once a year. It's for industry analysts. And I was there for, for a couple of days and we heard the dog and pony show on everything Workday's working on. I had a chance to talk to most of the execs, and they are also serious about AI and ml, but they have a very different perspective on it than Microsoft or Google. Their strategy is for the AI and ML to be used by Workday for Workday users on Workday data. So as many of you may know, back in 2014, Workday acquired a company called Identified, which was actually developing AI for employee profiles. And that eventually was used in the implementation of the Workday Skills Cloud, which is essentially an AI engine that infers and identifies skills.
Speaker 1 00:03:45 Remember, a skill is a word or series of words. It doesn't really know what your skills are in terms of how you use them. It just knows that there's some affiliation between you and these words. And the skills cloud is one of the things I want to talk about. So Workday's stated strategy is we have been doing this for a long time. The core system is built with machine learning in it. All the recommendations and forms that are filled out for you and analytics and places where are it pre-fills and predefines or pre recommends content or recommends job candidates. That's all ai. So Workday has machine learning engineers and has for quite a long time. But the big question about AI and Workday is a bigger architectural question. And that is ai. The fundamental difference between AI and traditional software engineering is that in traditional software engineering, you design an algorithm, you design a user experience, you build an algorithm, and you collect and analyze data from that system.
Speaker 1 00:04:47 That's how basically all of our computers have worked for years. AI is the complete opposite of that. It's inverted in the AI system. You develop algorithms, but the algorithms don't really know what they're gonna do until they get access to the data. And as they get access to the data, the system behaves in better and better and more unique ways. And the algorithms have a cost minimization curve that learns from decisions that do work and don't work. So the AI gets smarter by itself through its usage. So the key in some sense, fundamental differentiation of vendor A versus vendor B isn't the algorithms per se. I I think a lot of these neural networks use the same algorithms. It's the data, it's the quality of the, what is called the training set, which is the, uh, data that it's learning from the size of it, the nature of it, and so forth.
Speaker 1 00:05:44 In fact, open AI this week had a very interesting, controversial debate on Twitter about the fact that they are not going to release the training data. They're not gonna tell you what it is because that's their proprietary competitive advantage. And Google, of course, has a massive training data set they've been using in Lambda. So the reason I bring that up is that in the context of Workday, if you wanna build a big AI system about your human capital and your financials, a lot of that data is in Workday. Some of it is. So you could presumably start to intelligently understand who might be a better leader, who would be a good successor for this role, what is the financial impact of this reorganization so forth. But DA Workday doesn't really have that much data. The Workday customer base, which is huge. You know, there's 5,400 companies now using Workday, 50% of the Fortune 500 for Human capital.
Speaker 1 00:06:41 I think the financial application is growing even faster, although it's still a smaller market share, but there's about 60 million employee records in the Workday system and those don't belong to Workday. Those belong to customers. So in order for Workday to build cross customer ai, they need to get access to all of your data and much, much more data. Now, some percentage of those 60 million employee customer base have opted in to share their data with Workday's AI anonymously. So it's less than 60 million. And that data isn't that interesting. It, you know, the data in Workday is interesting obviously to the company and to the employee, but it doesn't have your job history. It doesn't have your LinkedIn profile, it doesn't have your work product, it doesn't know what software you developed. There's a lot of things it doesn't have. Let me contrast that system to some of the true AI platforms like Eightfold or Gloat or Seek Out or Sky Hive or some of the others.
Speaker 1 00:07:45 Those systems have a billion employee records in them and they are importing all sorts of data about people. Compensation history, job history, data, seek out looks at your GitHub submissions. If you're a software engineer, nursing data assessment data that has come from assessments you may have taken in different inside of your company. And one of them, I won't mention the name of them for example, can look at company A versus company B. And by looking at the job and profile information in the employees in company A versus the job and profile information and the job company B, and then looking at the people that are uh, maybe manager or senior manager and hire, IT can actually tell you the difference in leadership capabilities or leadership strategy or leadership culture between the two companies. That is a really, really fascinating thing. You couldn't possibly do that in Workday today.
Speaker 1 00:08:40 Now, I had this debate with Cheyenne and, and some of the other people there and their strategy to import all of this other data is to use Prism. Prism is the big data system that Workday acquired from Plate four and it was designed as an analytic system. I don't know if it can accommodate all of this unstructured data. So there's some limitations to what Workday's AI can do, but that's not, I'm not saying anything against it. It does a lot for Workday customers and it's gonna be ubiquitous part of the Workday suite As far as generative ai, there isn't a lot going on at Workday in that area at the moment. They have a few little prototypes are playing around, but they don't seem to have done much with it yet. But they're very aware of what it is and they're working on it. But let me add another big thing about Workday.
Speaker 1 00:09:26 Of all of the meetings I've been to at Workday, I must have been to a hundred meetings looking at Workday and talking to people. This was the first time that I felt like the HCM suite really came together in a very mature way. Workday HCM of course, is many considered to be the most advanced human capital e R P system on the market. You could debate that, but there's lots of things great about it. It's built on its own proprietary object model. It has its own global security system. It's highly scalable. It runs on multiple public clouds and the Workday private cloud. And it's very functional and very widely used in many industries. But in the early days of it, it was closed. It was, I used to think of it like the Apple iPhone. It was a beautiful system, but you couldn't do much other than what they gave you.
Speaker 1 00:10:16 Well, for many years they've been opening it up little by little, little and now they've really opened it up. Last summer, I went to the first Workday development conference and I was really impressed with the things they've done. And they have these two interfaces, one called Extend, which is APIs and uh, low-code development tools. It's not exactly for non-programmers, but it's not, um, as complicated as it would be if you just use the API itself. And they're making that easier and easier to use. And so you can build, there's, there's hundreds and hundreds of applications built by customers you have to pay for Extend, it's not free, but they'll train you to use it. And then this other tool called Orchestrate, which is, which is a workflow system to build complex workflows. So what Workday strategy is relative to AI and ML is to use Extend and Orchestrate and Prism to bring other data sources into Workday.
Speaker 1 00:11:05 So Fannie Mae for example, who was at the conference talked about how they're bringing a lot of their financial data into Workday. They're basically running Workday financials and their intention is to use the Prism system to do more intelligent analysis of the loans that they're giving and some of the characteristics of the loans. So there's a plan here and for the customers that have deep implementation of Workday and they have the HCM system up and running and they're using most of the modules inside of Workday, there's no reason why they wouldn't put more data into Prism and start to use AI against the whole dataset. I don't know if it will be as scalable as the other alternatives. You know, obviously you can do this in other systems. There are big data systems from aws. There's big data systems from Azure, Microsoft, Google, and they're all adding AI APIs right into their cloud offerings.
Speaker 1 00:11:58 So Workday's kind of going down that path, but they're clearly aware of this technology and they're working on it. And I think relative to the other ERPs, they're in really good shape. You know, I mean one could make an argument that the E R P idea of having a system for finance, a system for hr, a system for marketing, a system for supply chain, a system for manufacturing, system for patient care and so forth in different industries could really be disrupted by AI. Because a business AI system, which we will see somewhere in the next, you know, few years, should be able to look at data across all of those functional areas and give you intelligent information about your company's operations and how these different functions are working together. I actually know a vendor that's working on this and that's certainly where Workday's seeing the world going and they're trying to get there from here.
Speaker 1 00:12:51 I can't speak to Oracle and SAP's AI strategy at the moment, but I'll tell you more about it as soon as I talk to them. So, but getting back to Workday, hcm, Workday, HCM and Skills Cloud has really come a long way. The Core HCM system, of course has been around for 15 years. And over the years, Workday has added a lot of capabilities, the recruiting system, the employee experience, front end, this tool called Workday everywhere, which manifests Workday transactions through Microsoft teams, the Workday learning platform, there's a career hub. These things have gotten better and better and better. And we saw a demo, you know, originally, of course when they first come out, they're a little bit short on functionality because the Workday product engineers and managers have a lot to do, but they've really built it out. And David Sos, who runs the eight Sam M group, who I've worked with for a long time, has really brought this group of product managers together into a much, much more integrated suite.
Speaker 1 00:13:47 And we saw some demos that I think a lot of the specialized vendors would be surprised to see from Workday because what's happening in HCM from our perspective is what we call systemic hr, which means that none of the individual functions of HR can effectively operate alone anymore. You can't just hire people and optimize hiring without looking at internal mobility skills, data, possible development of people for these new roles, the job architecture, the retention and employee experience issues that might be preventing people from staying in these jobs. And even the job design itself and the job descriptions. So this thing that we used to think of as recruiting is much more complex. We call it recruit, retain, re-skill redesign. We call it the Four Rs. So the idea that the E R P or any other HCM system is gonna optimize one part of HR is great, but we needed to do more than that.
Speaker 1 00:14:45 We needed to interconnect. And so in some sense, where Workday's gone is really where the HR function has to go into a more integrated way. Now that is not to say that the standalone HCM vendors are about to be crushed by Workday. That's not true. We work with all of them. If you look at Eightfold and the second and third order AI and matching and intelligence capabilities they have, those are things Workday can't do. If you look at Beamery and the types of sourcing and intelligent selection and candidate relationships, they can do Workdays not at that level of capability. If you look at what Cornerstone and Degreed and Docebo can do in learning, there's features that are missing in Workday and you can go through gloat and there's things in Gloat that Workday hasn't even really thought about in the talent marketplace. So virtually every part of the HCM suite is represented in Workday, but there are still really incredibly strong other vendors in these niche niche or, and I wouldn't call 'em Niche, but in these other functionaries that are very, very good also.
Speaker 1 00:15:48 So I don't think there'll be a lot of consolidation. However, one more thing, and I talked with Anil about this. You know, there is a recession, there is a slowdown, there is going to be tighter budgets. Most of you aren't gonna be have unlimited budgets for HR technology and somebody in your company and IT or HR is gonna look at all the stuff you're buying and say, how come we have all these tools? Can't we put it all in Workday? And you're gonna have to defend that. And so I do think for the next year or two, until we come out of this economic malaise, we're going to see more consolidation and workday's in a very, very good spot because of the integration and the functional sophistication they've added. Let me talk about the Skills cloud for a minute and then I'll cut this off cause I don't wanna go too long.
Speaker 1 00:16:35 So everybody that I talked to that's in a Workday environment is asking questions about the skills cloud, what should we use it for and how do we get it up and running? And I think that's the wrong idea. The Workday Skills Cloud was originally conceived to be a skills engine for the Workday system. And during the time it was first developed, there weren't a lot of skills engines out there, there wasn't a lot of conception of how this was gonna work. The AI models were a little bit new. We didn't have chat G P T. So it was really a very simple system. It was really a database for skills. It did some first generation AI to de try to determine what an employee's skills were by looking at their performance appraisals and their profiles and things like that, their jobs, their job titles, job descriptions and so forth.
Speaker 1 00:17:25 And people kind of were interested in it and they weren't sure what they were gonna do with it, but they thought it was a good idea. So most people licensed it, most Workday people licensed it. And all of a sudden there was this huge barrage of skills technology from other vendors, the greed and now Cornerstone and EdCast in the learning area, eightfold and iSims and Beamer and Phenom in the recruiting area, gloat and Fuel 50 and a bunch of smaller companies in the tele mobility area, the large skills technology vendors like Sky Hive and Retrain and others. And these companies were much, much more advanced in their thinking than Workday at the time. They were building skills engines from billions of records with advanced ai. And I think Workday discovered that they were not really there yet. And so many of the Workday customers believed they could use the Workday skills cloud as the repository for all these skills.
Speaker 1 00:18:22 But in reality, as you may or may not understand, the skills dialogue or the skills issue is not like competencies. These are not things you write down and design. There are tens of thousands of skills being created all the time. Every time a new technology or algorithm or tool is created, there's a skill. So the skills technology that we're coming to build is really a living, breathing, metadata management system. I like to call skills as the metadata of humans. Really everything about you in some sense could be called a skill. So what the Skills Cloud, you know is becoming, is really a skills platform that can consolidate and centralize a lot of the information that's being used for particularly specialized purposes in the recruiting tools and the tele mobility tools in the PAT talent marketplace, in the learning system, even the pay systems, by the way, if you read the research we just put out this week on pay equity, there's no way to fix the pay equity problem in companies until we understand skills and refer pay to skills too.
Speaker 1 00:19:27 So the Skills cloud is kind of taking on a new life and it's come a long ways. So for those of you that are Workday customers, I think you'll be surprised that it is not as difficult to understand as it used to be. And they're doing a lot of things to make it, uh, more functional to use. Now that doesn't mean that the skills effort in your company is a technology project. It is not. And we do a lot of skills workshops for companies and we're willing to do this for any one of you. You need to think about skills as a means to an end, not an end Skills is a technology that can help you improve performance retention, recruiting, leadership development, succession management, et cetera. Sitting around and debating what your skills are and putting them into a database doesn't actually solve those problems.
Speaker 1 00:20:18 So we really encourage people to start with the problems and the domains and look at skills and capabilities and academy formats before you go out and try to build this massive database. Because whatever the size or quality of the skills data you get from Workday or anybody else, it's gonna change one of the vendors. That's a really interesting vendor I want to encourage also you to think about in the skills area is this company called Light Cast. Light Cast is the merger of MZ and Burning Glass. And they have, as an independent data provider, they have taken all the data from the O net database from the federal government and built their own skills ontology that is openly available through an api and they have no vested interest in you buying any particular software. And they're licensing the open skills architecture to more and more software vendors all the time.
Speaker 1 00:21:07 So if you're looking for a standard taxonomy of skills, take a, take a look at Light Cast, we can put you in touch with the right people there because they're one of the vendors that is maybe the only one that's trying to build a completely open, neutral, licensed complete skills architecture and job architecture based on the data they get from tens of millions of job postings every day. So that is a source of architectural information to help your skills strategy feel more standardized. But I am pretty high on Workday at the moment. I think they're really in a good product cycle and I think you're gonna be pretty excited about all the things they're doing. Okay, a couple more quick things from me on what we're up to. We have launched the beginning of the website of the irresistible conference June 20th through 22nd in Los Angeles.
Speaker 1 00:21:58 There's a whole bunch of stuff not on the website yet. We're gonna have a very special guest to talk about ai. We're gonna have a panels of CHROs talking about strategy, retention, employee experience, people sustainability, ESG diversity. We have all those really amazing experts coming. There are some development programs there for your teams. If you're a do person, company, corporate member, you're gonna have a very special event. You're gonna visit some clients down in Los Angeles and see some things that many people have never seen before. And on Friday we have a special irresistible education event for the members that come to do a simulation on your own irresistible capabilities and tell you a little bit more about that in the next couple of weeks. So that's now out there and we really encourage you to come. Last year the conference filled up, it's undoubtedly gonna fill up again, it's at usc, we can only support 450 people.
Speaker 1 00:22:54 So bring your team sign up. There are not a lot of vendors coming. There's some vendors sponsors, but we don't have vendors wandering around the conference. It's really for you. It's for HR people and business people. The other thing I'll just highlight, we introduced our pay equity research. A lot of you may have seen that it's very, very significant topic because pay equity impacts everything that goes on in business. We had a meeting on Friday with a lot of you to talk about the HR transformations going on in the work that we're doing on the operating model. And one of the things that came up in many of the conversations was if we're gonna have a more agile HR function, we're gonna have to have a more agile and fair reward system because people don't want to work on new projects if they don't know what it's gonna do for their career.
Speaker 1 00:23:38 So anyway, take a look at the pay equity content. Soon after this sometime in May, we're gonna launch a very significant piece of research on the whole rewards strategy for companies. And that's gonna be another eye-opener for you. And let me mention the Josh Person Academy. The JBA is growing like mad. We have a new course in there on rewards and you're gonna see a lot of new content. We've really doubled down on the content going into the jba, so I'm going to spend more time on it in the next couple of weeks and tell you more about what's happening there. Anyway, so that's a little update on AI and Workday. We are spending a lot of time on the AI stuff, so I'll be continuing to tell you more about what's going on in generative AI and some of the amazing things going on in the HR domain in the next couple of weeks. Okay, thanks everybody. Have a great weekend and have a great week.
