Episode Transcript
[00:00:00] Good morning. This week is a pretty major HR technology conference in Las Vegas here in the us The Unleash conference. So there will be a lot of product announcements and technologies demonstrated and discussed here, including some from us. And so I spent some time over the weekend writing a fairly high level article on where this is all going. So I'll send a link to it. But the general perspective I want to explain is how different in technology in architecture and use case HR technology is going to be. You know, as somebody who's been doing this for a long time and spent a lot of time back at the old mainframe days, I think if I give you a little perspective, it'll help you make sense of all these tools and agents and AI things that are suddenly being thrust upon you. So if you sort of Fast forward backwards 20 or 30 years, HR technology was payroll and record keeping and it was used by bureaucrats and administrators and HR people and salary administrators and payroll administrators. And we never saw it and we never knew about it. I remember in my IBM days when I would go to get a performance appraisal, my boss would pull out a piece of paper that I had never seen that had all this information about my job history at IBM and then he would give me a rating 1, 2, 3, 4, 5. A lot of times it was a 3, which always irritated me and talk about my development plan and send me off back to work. And I didn't know what was going on in the computer about me and it was a little bit creepy, but I didn't really care because I had other things to do in the ensuing 15 or 20 years. All that mainframe stuff, much of which was custom coded by the company, by the way, there weren't too many application vendors at the time. There were some turned into a client server, PC based and then cloud based systems again for HR. And in the early days of PeopleSoft and other related companies, this was a component of the ERP system. And the idea was that the human capital or HR stuff would manage the people assets the same way it would manage the physical and financial assets. I mean essentially there's three kinds of assets in a company. There's the money, the people and the goods and services that are physical and the inventory and all that.
[00:02:24] But, but as you know, there's much, much more than that in reality. There's all the intellectual capital and brand and business processes and customer value and you know, et cetera. But anyway, so those early ERP ish things took the form of PeopleSoft and Oracle and others that were building them. And the technology vendors kind of got sucked into that by selling databases and middleware. There was all sorts of middleware that was trying to stitch the various heterogeneous databases together. And the software industry built more integrated Systems. You know, PeopleSoft was one, was the first that really understood the relationship between all the human capital practices. And by the way, in the middle of all this, talent management emerged as a bunch of software products, applicant tracking, learning management, onboarding, management of payroll and pay equity, and DEI monitoring and all sorts of things there. And so there were integrated talent management systems, integrated recruiting systems, integrated employee engagement tools in the big hcm, but it still wasn't really used by employees much, a little. And then what happened along the way is the concept of an employee portal or an employee experience location where you could go find things you could look up not only what was your 401k balance and your vacation balance, things like that, but you could look at your learning, your development plans, you could find content to do work and to improve yourself and to and so forth. And then a bunch of tools were built by many companies, including the big ERPs, to expose this data to employees. And it was sort of a little bit fleeting in the beginning, but it became very big as Microsoft got into it with VIVA and many other tools. Companies, some of them relatively small, built these beautiful experiences for employees. But the bulk of the engineering and the system design was not designed for people, it was designed for the business and for administration.
[00:04:26] So you had to build a lot of layers of analytics and logic on top of that so that employees could understand what it was doing for them. Even something like performance appraisal and performance management never really got built out very well by the ERPs. So there were a whole bunch of vendors that did that. So we've got this kind of layer cake of very complex, highly industrial back office stuff with a little bit of a rat's nest of tools on top of it, making it easy for us to find. And if you're a frontline worker and you're scheduling your shift or shifting, you know, asking for a day off, this stuff's important.
[00:05:07] So what happened is a lot of the users of these systems got frustrated, you know, so SAP and Oracle and Workday were not very good at making the front end part because it was hard and it was new and the systems weren't designed for employees. So companies got frustrated. They built their own custom portals. A lot of companies built front ends and didn't let the end employees touched the backend system. At all. They built something on top of it and there was a lot of money spent on that. I mean, I talked to one big pharmaceutical company that told me they were spending a hundred million a year on their employee portal just managing it. So anyway, along comes the mobile phones and then we build mobile tools that were easier to use. And the mobile applications from these vendors were great and they were much, much more useful and kind of broke this lockstep problem.
[00:05:56] And then we kind of enter the world of AI. And by the way, while all this is going on, desktop productivity tools are exploding. We have the pandemic, we have Zoom, we have Microsoft, we have Google. And a whole, even, even Meta got into this for a little while, building employee communication tools, conferencing tools, email systems, note taking systems, knowledge management systems, training stuff. And, and so there was this massive industry building, not really that well connected to the back office software at all. It didn't really matter because all you needed to do is, you know, get a little bit of data from it. And those companies grew to be very, very large because we were very, very focused on using our productivity applications, our phones, our computers and ever. And anyway, so this back office stuff is sort of, you know, trying to reinvent itself. Now fast forward to today, 2026.
[00:06:48] Now you have AI. So we all know that by simply going to ChatGPT or Gemini or Anthropic or Galileo, you can ask a question and within 10 seconds get an amazing answer. Now that data doesn't come from the internal systems, it comes from the LLM itself. So the AI companies, the frontier vendors built data management infrastructure that was wicked fast. In fact, you know, I, I'm always sort of amazed and particularly with ChatGPT, it almost answers the question before you press enter. I don't know how it does it, but I think it's because it's simplified and cleaned up all of that data in a form that the vector indexes are sub millisecond when you ask a question. Now in the corporate world that's impossible because the data is located in all these back office systems. There's 14 or 15 or 20 different employee systems behind that and they're slow to respond. And the APIs are old and the technologies are old, so, so you can't get it that fast. So it's not as simple as just sticking anthropic on top of all your infrastructure. There's a bunch of middleware and agents and piece parts that have to be built in the middle. But there's no question that the Market and the demand and the use cases are moving away from the back office towards the front office. Now there's still a lot of hugely important back office things that need to be done in analyzing the org structure and looking at skills and looking at turnover and looking at productivity, revenue per X, Y, Z and you know, many, many things. Automating the process of recruiting, automating the process of career, automating the process of project management. Lots and lots of things to do on the back end too. But again the systems weren't designed for that either. So now we've got these multibillion dollar HCM companies and there's a lot of them. It's not just Oracle and Workday. I mean there's, there's probably 50 of them when you add them all up, including smaller companies that are not that small anymore. Bamboo hibob, I mean they're all billion dollar market cap companies are more UKG dayforce, on and on and on. They're pretty big. So they've got this kind of slightly legacy back end and then they're building all this stuff to make it easier to use. And so the AI is eating into this market from the top and trying to crawl through all these backend workflows that were built, you know, maybe 10 years or 20 years ago. Somebody may have designed some of this stuff a long time ago and that engineer may have retired. So there's a lot of re engineering and redeployment happening. And the challenge for us as buyers and users of this is how do we navigate this space? Do we assume that the incumbent vendors we have will eventually build this beautiful front end? Or do we ignore that potential noise and look for somebody who's focused on it? Along comes ServiceNow, who is a very big company. They're two or three times the size of workday and they basically say, you know what? It's all about the workflow, it's all about the employee, it's all about the agents. The backend is legacy. You can, let's assume it's just there, let's build something on top of it. Let's give you the development tools, let's give you the security infrastructure, let's give you the APIs pre built and they've been really successful selling that. Now the problem that ServiceNow's had is because they're run by a very sophisticated sales oriented executive is they've really sold a lot of expensive software that companies haven't deployed yet. So it's a very elegant system with lots and lots and lots of features and capabilities. Most of the companies I talk to say they have it and they spend a lot of money on it and they're not using it that much yet. So the question is how do you make it easy for IT departments? Or maybe it isn't even it, maybe it's the employees themselves to get a connectivity layer built so they can do their jobs better. And that's what this article is about when you read it, is I'm walking through the future of this and what it's going to look like. And rather than you as HR people or business people kind of mapping this out on whiteboards, I think it's actually useful to take just an even higher level view of what would you like the system to do? As you'll read about in the article, we would like the system to know who you are, understand what you're trying to do, interpret your question or your request, go find the systems and data it needs and do it for you. So if you're a manager and you have a performance problem on your team, you would want the system to understand the question you're asking, get the data it needs to interpret the possible situation and give you a series of recommendations. That's a multi layered problem. The HCM vendors don't really know how to do that because they don't think all day about high level business problems. They do understand them, sure, of course they do, but they don't engineer solutions for that. They engineer solutions for the technology back end for the most part. So we need a techno an intelligent orchestrator we call Galileo an intelligent, an intelligence orchestrator because that's what Galileo is. It knows how to answer these questions, just doesn't know how to find the data yet, but it will. And then the middle stuff in the middle from Microsoft or Zoom or Google or ServiceNow facilitates this agentic relationship. And you can certainly see a world not too far from now. In fact it's already here where an employee goes to an agent and UKG demonstrated this at their conference and says I need about $500 of extra pay for Christmas gifts from my family. What shifts am I qualified for in the next three weeks to make an extra $500 after tax. Now that means the system has to find the shifts, find look at your skills relative to the shifts available, determine how much extra money you're going to make, understand your tax rate and come back and give you an answer. So there's like a whole bunch of back end things that it needs to look at in order to answer that question. But they actually showed a demo of that. I'm not sure it exists, but it's probably the. But it's the right idea. So the world of HR tech is not about HR anymore. I mean a little bit of it is, some of it is, but, but it's much more about employee productivity, employee engagement, employee wellness, employee support, employee experience. And so using the miracles of AI and the analytical and data management capabilities of these vector based data systems, that's where we're going. And the reason that I think it's important to think about that is when you buy the piece parts in the middle, the ServiceNows or SANA or Workday or whatever you know, comes along and there's a lot of them out there. Viva from Microsoft. I mean all the learning companies are getting into this. It feels like a relief when you do away with a problem you have and say, oh finally people can find this thing they've been looking for. But then all of a sudden they say, well that's not enough. I don't want to just find it, I want to ask it to do something for me. So you have to think about this from an architectural standpoint. Now I know a lot of the vendors in the market really well and we work with many of them because we're analysts and because we tend to talk about how Galileo interacts with them and stuff. And there are very few able yet to do this. But I think what's going to happen, and I think the big vendors are beginning to kind of come to grips with this is a big collision course between the technology companies that are really good at user stuff and the technology companies that are really good at business backend stuff. By user stuff, I mean Microsoft, Google, Apple are exceedingly good at understanding how to deliver something to you that's easy to use. Oracle, SAP, Workday, ADP are very good at understanding the backend processes that have to happen for the company. And in the middle is a bunch of money to be captured by these companies.
[00:14:59] And what you're going to see at the conference, and I will talk much more about this after all these announcements come out, is the attempts by the existing HR and ERP technology vendors to become the front end. And at the same time the expectations that we have using the tools we have today on how well these things work. And I'm going to sort of predict that there's going to be a lot of shakeup in this business. The primary reason for the shakeup has to do with money. You know, if you buy Oracle, SAP Workday ADP. Hi Bob. Whatever. And you pay $50 per employee per year, $75 per employee per year, $100 per employee per year, whatever it turns out to be. And you add that up, it's a big number. And the IT department has budget for that because they have to, they just need these systems. They're never going to go away that you have to have a system like this. But then the LLM vendor comes along and the application productivity vendor comes along and says no, it's, it's $20 a month for my thing, it's $30 a month for the copilot, $30, $200 a month for cowork. Somebody with a green eye shade on looks at that and says no, we're not going to spend that much money on everything here unless we see a huge roi, unless we see know, massive improvements in productivity. So somebody's going to compete for that new budget that we spend on AI. You know, I looked last year and I'm going to be talking about this at the keynote. The IT spend on AI is massive. IT budgets, according to the report I read this last couple of weeks are up like 63, 64, 65% on AI infrastructure, AI capacity and so forth. That's going into companies, businesses as well as tools. And most companies are not sure how well they're being used yet because it takes time to absorb them and find the right use cases. So if the HR tech vendor says we want an extra $30 a month per user, whatever it is for their stuff, somebody in IT and finance is going to say, well hang on a minute, we just bought Gemini and Copilot and Anthropic for the engineers and a bunch of other stuff. Are you sure we need that too? So you could expect, I mean you could predict if you were a financial guy or gal, that maybe the incumbent SaaS companies get acquired by the AI companies. Because you know, the funny thing about the market right now is the AI company's market valuations are 20, 25 times sales and the valuations of the software companies are four times sales, five times sales. So you, you as an AI company could buy an ERP if you wanted it. I mean, I don't think they do. I think they want nothing to do with it, but. And you could just deliver it and re engineer it and take all that revenue and use it to rebuild those tools. So it's a very disruptive time from a marketplace. So this is more food for thought today. I will come back later in the week as all the announcements come out and give you perspectives on the reality of what's happened. I'll link you to the article I wrote over the weekend. It's not very long and I think you'll find it informative and a little bit thought provoking.
[00:18:18] And we're here to help you through this. It's a very exciting time. It's all moving in the right direction, but you know what happens. There's a lot of shakeout in the process. Have a great week, and those of you that are in Vegas, I'll see you at Unleashed.
