Episode Transcript
[00:00:00] Okay, I want to talk about why the HCM market is likely to be a very fast growing space again. Now, fundamentally, HCM systems, which are oftentimes considered systems of record or payroll systems, have been around a long time and in the early 2000s they all went to the cloud. And since the last 25 years, since then, most vendors have built out talent management tools and various employee experience tools into these platforms with some degree of success, to the degree that gartner estimates around 9% of all IT software spending is in HCM. And that sort of makes sense to me because managing people is a pretty big part of managing a company, but there are many other things to do. So 9% feels roughly appropriate. And they forecast a year or two ago there was about 70, 65 to $70 billion.
[00:00:56] And the way I look at it is there's submarkets around the HCM market, there's the applicant tracking, recruiting software, there's the learning software, there's software for benefits management, there's software for employee experience, there's portals, there's case management tools, all sorts of stuff. But this core group is important because you need that to run a company. You to need, you don't need the other ones necessarily.
[00:01:22] And if the core system has some of the capabilities you need, you may not decide to buy some of the other add ons.
[00:01:29] Now, as I talk about in detail in the article I published today, a fundamental shift is taking place. The argument against these HCM platforms in the world of AI is that they're just dumb databases and you can build AI agents on top of them. Well, that's a really naive perspective and I'm really sort of shocked that so many technologists say this because you can't just build an application on top of your HR system without being very careful to protect the data, use the data appropriately and not update the data randomly because you'll mess up your company.
[00:02:08] And the processing of payroll is very complex. So if the data is incorrect on how many hours you worked or your shift or your level or your pay or your Benefits or your 401k, it gets really messed up in a hurry. So we really can't have third party applications chewing away at these HCM platforms unless they're done very carefully. So the HCM vendors over the years built APIs and gated interfaces so that when people went in there, they mostly did reading, didn't do a lot of writing. The systems of record still stayed as the mothership of, you know, kind of like who's working for your company. And how much they're making and what they're doing. And if you think about the frontline workforce where 72% of the employees live, their schedules and their skills and their certifications are very important to their jobs. So there's aspects to this that are very operational in nature. Not just every two weeks, cut a check. Now move into the current world of AI and just bear with me and assume that these agents that we're going to either build or buy are, are going to be doing things like people, but even more complex because they can manipulate vast amounts of data at high rates of speed. So they will be assessing job candidates, interviewing job candidates, scoring people, screening people, identifying people's skills, giving people development plans, monitoring their performance, training them, coaching them, giving them recommendations on their benefits, helping them schedule their day, helping them schedule their shifts, helping them get promoted, helping them decide whether they need a new position inside of the company. Search out internal candidates. There's a lot of, there's, we have 240 use cases for this and these are things that get done by hand today. And more and more of it's going to get done by machine. So it's going to get done intelligently with good data. Where's that data coming from? Some of it comes from the individual human being, but a lot of it comes from the hcm and more and more of it will. So you can't build applications that do all the stuff I just said without using the data data in the hcm. So the question that you might ask yourself is why wouldn't you use the agents from the HCM vendor to do this first, assuming that they build good stuff and that they give you a nice experience and a development tool to build your own applications on top of it? Well, that's exactly what's happened. And as you read about in the article, the new world of HCM is very, very different. It isn't just a dumb system of record or a payroll application. It is a system of productivity growth and work. Now, this isn't all built out yet, but it's very clear this is where it's going. And that means the value of the HCM software is about to go up a lot.
[00:05:02] And those financial analysts that keep beating up on these enterprise software companies are going to eat their words, in my opinion. Now, in the article, I discussed six areas of transformation where the HCM market is moving in, including that these platforms are now very, very easy to use. They are now development platforms. They now have benchmarking and other forms of third party data. In them, they have management applications in them. And they are designed to monitor and manage and schedule and optimize the work of humans and agents. Those are operational things that we want to do in a company. So if you're reorganizing sales or reorganizing manufacturing or buying a company, or growing like crazy or shrinking like Microsoft just did yesterday, somebody's got to make a whole bunch of decisions about who's going to do what and what jobs they're going to do and what agents are going to use. And if that gets done by hand, it's not going to get done super accurately or very quickly. But the AI agents in the HCM can do that. So my thesis, as you'll read about in that article, is that these platforms are being redefined as systems of growth and systems of work. And I know specifically what's going on at SAP. I know what's going on at Workday, I know what's going on at hibob, and I know they are doing some very unique things to not only attach the HR platform to the finance system, which is a relatively small part of this actually, but to build a work management experience on top of of the hcm. Now, financially, what this turns into is you as a buyer or a consumer or a consultant are going to see new revenue or new cost to run these systems, because these companies are going to charge you buy tokens, in other words, buy activity. So instead of just licensing you a bunch of agents and hoping you know how to use them, the agents will probably be free or close to free. And then as you use them, you'll pay for consumption. And if the agent doesn't work or it doesn't do what you want, you just won't use it or you won't pay for it. So the token expense is incremental, not replacing the licensing fees you're paying today. Now, you know, you might say to yourself, well, workday is already too expensive, or SAP is too expensive, or why are we spending all this money? I don't think that really is the right question. Because if HCM is only 9% of enterprise software today and your labor costs are probably 50 to 60% or more in some companies, it's much higher in healthcare, for example. Then why wouldn't you have a system that really optimizes that labor and digital labor? Well, I mean, why wouldn't you pay 15% of your IT cost on that? I'm not sure it'll go that high, but it could go higher and it'll be worth it and I know this because we've done case studies in the recruiting space. If you implement a tool like Paradox, for example, which is now owned by Workday or Maki People or another modern AI driven recruiting tool, and you can hire people into your restaurants faster, or UKG for example, and you can staff your hospitals more accurately or you can staff your truck drivers more quickly, your company is going to grow faster and operate at a higher rate of speed, generating more revenue and profit for your shareholders. So that system now is becoming a system of financial performance, not a record keeping HR system in the basement. And this is happening. This is not a dream that I just made up. This is real. In the area of learning and development.
[00:08:43] When you redeploy people into a new job or you launch a new product or you have a major crisis or something changes in the regulatory landscape, the old way of dealing with that was you create a bunch of courses and you force people to go through this training process and maybe in three to six months everybody kind of gets it. Well, in the new world, the HCM dynamically sees this change and gives everybody in the company, within maybe days or hours, an update on what they need to know and lets them personally learn as they needed to deal with the changes that took place. Because sometimes they might be safety related or compliance related or other things. So these are spectacular business performance features that never existed before.
[00:09:25] And one of the other reasons the HCM system was never perceived this way, it was too hard to use. I mean, I know a lot of companies, even Google, that basically took Workday, which was the easiest to use of all of them, and hid it in the basement because they didn't want people to have to learn how to use it because it was too complicated. Not that it wasn't designed well, but it just has a lot of features designed for HR and it people that end users don't really need to see. Well, the new user interfaces, Sana from Workday, Jewel Work from SAP, the new Redwood interface from Oracle, the new hibob Agent from hibob and others from ukg. The bright system from UKG are so easy to use, you don't need to know anything about the system. You can literally talk to them or ask them questions and get the answers. And the development tools in these platforms are as good as these other off the shelf development tools in the market, like Lovable and others. In fact, I think the flurry of excitement about development tools is over. Elon bought one of the big ones, Cursor, and the development tools coming out of Workday and SAP and other vendors are going to be just as good as the development tools you have from Claude or anybody else. So why wouldn't you develop there if the token costs are appropriate where it's already integrated with all the data and security you have Now I run a company, I know what it's like. We already have too many systems in our company. If I could use hibob, which we use for all of this work related stuff that we do, I would do it in a minute. Just to know that it's all in one place and it's all integrated and even if it cost a tiny bit more, it's worth it. So from the standpoint of the revenues and market cap of these HCM companies, I think they're going to go up. Now as you'll read about in the article, the Workday strategy and the SAP strategy and the other vendors are all slightly different. Not everybody's going after this market the same way. SAP has a much more complex environment to deal with because they have just more functionality in different areas than Workday does. And then the smaller vendors like hibob, ukg, Dayforce have, you know, smaller resources to do this. But this is definitely coming and in a sense it's a recentralization of where IT money is spent. Just to give you a sense of the IT spend by the way, and I put some Gartner data in the article, the Gartner guys believe that the IT software and infrastructure market for corporate will be 7 trillion next year, grow 15% and that the consumption token part of it is going to be 1.4 trillion and grow at 56%. So this consumption based spending is big and it's growing very fast. So you could see, and we'll see what happens in the next couple of quarters. You could see Workday, SAP, other ATMs generating good size double digit growth from the consumption of these agents as companies start to use them. Now there's a couple of other issues as I talk about in the article. First is ServiceNow. ServiceNow has done an amazing job of building infrastructure around these things with development tools, case management tools, knowledge management tools, infrastructure tools, agent management tools to sort of prevent you in a way from getting locked in. And that's always been an attractive sales message that you get locked in. I mean, I don't know if I mind being locked into one platform if it's good. Frankly, you know, I'm locked into Google for search and it doesn't bother me that much. But they've been Very successful at staying ahead of this market and preventing the corporate buyers from trying to build everything on top of their HR system, because nobody has one HR system. First of all, most companies have multiple payroll systems and then they have learning and recruiting and all these other various applications. So they're not going away. And they've done some amazing things as well. So they're a player here. And I think some of you are going to be ServiceNow shops and you're going to be debating about should we build this in ServiceNow or should we build this in Workday or should we build this in SAP and so forth, depending on the architecture and what the application is you're trying to build. The second issue is the smaller vendors. And for most of my career I've been big fans of innovators, new tools for coaching, for training, for employee experience, for work, life balance, for scheduling. And you know, there's thousands of really innovative startups building cool stuff. I think they're going to have a little bit of a harder time. The big vendors do want to have ecosystems of partners, of course, that have agents that plug in and they don't want to be black boxes. So they've opened up their APIs now, as you'll see in the article, to make it much easier. But if you get a coaching tool from BetterUp or Valence or another company that's specialized and then you get one from Workday, the Workday one can access much more data in the company than the third party one. And so if Workday put their energies into it, which they are right now, it might end up being a better system just because it's integrated. So there's going to be a little bit of a shakeout in terms of the highly successful third parties interacting with these big HCM platforms. Now that the HCM platforms are pretty powerful and pretty good development tools in their own right. The other big question out there is the Microsoft Copilot. And I do believe that Unlike Claude and ChatGPT, the copilot will and likely be a major development platform for HR stuff because we all, 70% of the world has Microsoft tools sitting on their PCs or their phones. So if you are an IT person or an HR IT person and you were trying to build an application for, for coaching, onboarding, employee record keeping, scheduling, you would want to make sure that it works pretty well with the desktop systems that you have. And so we will see what happens. Relative to Microsoft, the Microsoft tool set is really easy to use. I mean you can build an agent in Microsoft Copilot by just talking to it. But you can do the same thing in Workday and you can do the same thing in SAP now with Jewel work. So I think there's going to be a sort of a whole bunch of experimentation on where's the Copilot better versus where is the HCM better for employee apps? And you know, the thing about employee and HCM apps is they're not standalone apps anymore, they're part of a network. And what we really have been working on in HR 2030 is showing you how these agents work together. And so if you end up building a couple of agents in Claude and a couple of agents in Copilot and then a bunch of agents in SANA and then a bunch of agents in something else, it may not be that easy for them to communicate with each other. And you might actually be stuck in a little bit of a bag of doorknobs, I used to call it situation.
[00:16:18] Especially when the world is so new at this and we don't have good architectural blueprints yet. So the jury's a little bit out there. And I know Microsoft has huge investments and plans and expectations for this. And those of you that are Microsoft shops are probably going to use the Copilot just because you're, you just understand the Microsoft infrastructure because the Microsoft Graph data, the email data, the document data, the meeting recordings, is not really the same as the HCM data. It's different. It's in a different place in a different structure. Now, interestingly enough, SANA from Workday has a lot of the features of the Microsoft Graph. It has a meeting recorder, it has the ability to read emails and import emails into Sana Enterprise. And I'm sure SAP will figure out how to do this at some point. I don't think they're doing it yet. Yet. So you could, if you're a workday shop, possibly integrate your Microsoft environment right into Sauna and may not need to build a bunch of apps in the Copilot that don't necessarily sit directly in the workday experience. But that's kind of a company by company decision. We, for example, because we've been using SANA for about four years, have a lot of Microsoft stuff and a lot of Sauna stuff. And I would say more and more of our functionality is moving into Sauna just because it's a little bit more integrated experience. But not everything is there. I mean, the Microsoft documents and SharePoint environment is very, very big. And we have just as much investment as that as most companies. So that's another one to consider. The other topic that I think is important here that is not discussed in the article, but it's discussed in our meetings with these companies is the architecture of the agents themselves. If you compare the agents that are now on the street from workday versus SAP versus Oracle, what you see is Oracle has a thousand. So I think to Oracle, any automation that automates a tool or utility or data system is probably considered an agent. SAP has hundreds which are more like functional specialists and workday has about 17 or 18. So in a sense workday is building super agents which are functional business agents that have sub agents beneath them. The SAP guys are building automation agents that are grouped together into what are called assistants in Joule and Oracle, from what I can tell is just building agents as fast as they possibly can. You're going to want to know how these things work together because the power of agentic HR is the interconnectedness and the systemic nature of these agents working together to coordinate and share information and learn. What you want to have happening over the next couple of years is you want your AI to learn things about your company so you get smarter about who to hire, how to train somebody, what a high performer looks like when an agent is better than a human, when a human's better than an agent, what is the leadership model that's working versus not working, what are the behaviors that are solving customer problems and in creating customer value and what are the behaviors that aren't and so forth, those critical human capital, somewhat scientific but non deterministic things. We can get great insights from AI there. So if the agents aren't interoperating with each other, well that may not happen. And it's not clear to me in the case of SAP and Oracle if they're thinking that way yet. They will of course, but if they built a bunch of standalone automations, they may not have considered how they talk to each other. What we did in HR 2030 the reason I'm sensitive to this is what we did in HR 30, because we were doing this conceptually, is we already thought through the kinds of interoperability and interactions that will take place between these specialized AIs and tried to diagram out the workflows and the data flows that will happen when a given one of these 240 business cases takes place. Now we're not software people, so we're not going to build it, but we can see it and explain it. And for those of you that use any of these big platforms, that kind of architectural view will help you plan what you're buying and building and also help you interact with the vendor to coach them along the way as to what you would like them to build or what you build yourself to interconnect these things the way you want them to work. So another big part of this re energized HCM market is you have much more control now. You don't have to wait for the vendor to put out the next release to get something fixed. I mean, it amazes me how much waiting there is for features. And there's only so many product managers in one of these companies. If you look at the million businesses around the world, each one of those businesses has someone who's kind of an IT person who could be building something on their own. None of these vendors have a million product managers. No way. So rather than now waiting for the new version of XYZ to come out from vendor abc, in the new world, you're going to be able to say, you know what? I don't even know if they're going to build this. Let's just build it ourselves and get it working. And then if they announce something, we'll turn ours off and use theirs. And that's a level of flexibility and value that you never had before too. So big, big, big, big, big topic, lots of debate we could have about this architecturally and competitively between the vendors. Read the article. I'm interested in your feedback. Come to one of our HR 2030 roundtables if you'd like to discuss what's going on in your company. We'll be happy to help you with this. It's a very exciting new world. And stay tuned for much more to come in the HR 2030 experience. Bye for now.
