Episode Transcript
[00:00:00] Good morning everyone. Today I have a really interesting, important announcement to discuss. I happen to be in London this week and have had a lot of meetings with mostly financial services companies about their AI transformations.
[00:00:15] And so later I'm going to explain a little bit about what's going on there. But let's talk about Workday and Paradox. There's a whole bunch of dimensions to this acquisition, all of which have significant implications on HR tech and hr. First of all, I think it's a really intelligent strategic move for Workday for two really three reasons. First, the talent acquisition suite at Workday definitely needs help and this adds a significant enhancement, which I'll explain. Second, this has the potential to advance Workday's architecture towards a more AI centric approach, which it is not really today. So far they're working on it. And third, I think they get a pretty incredible team and some business approach and not only thought leadership, but operational leadership that I think is extraordinary in Paradox. It could help Workday a lot if they choose to adopt it. So first of all, who is Paradox? Around seven, eight years ago there was a, before the AI craze, there was a flurry of interest in chatbots and a lot of companies, vendors started to build tools to mostly based on NFTs LP to deliver chatbots to job candidates. Almost all of them went out of business or acquired because they didn't really have the underlying architecture to do more than chat Paradox. Whoever was forward thinking enough to realize that what they were really building was a conversational interface to the entire talent acquisition process. I don't know if they knew that in the beginning, but they figured it out pretty fast. And so what they ended up building over the years was, was a chatbot originally called Olivia with the things that a job candidate would ask, you know, what are the hours, pay levels, skills required, location for this job, Mostly for what was called frontline work. But later it expanded and then they realized, well, we should have a way for this candidate to apply, we should have a way for this candidate to go through the interviewing process, we should have a way for this candidate to be assessed, we should have a way for this recruiter to look through the pipeline, we should have a way for the candidate to do onboarding, we should, after they join the company, we should have an employee experience also based on the chatbot. And so what they were basically building was an end to end recruiting system based on AI. It wasn't called that at the time, based on conversational interfaces that could scale and deliver at incredible speeds that you could never do in a traditional recruiting process. Because in a traditional recruiting process it's a stove piped, handcrafted, step by step, non integrated system. Find a job that's open, write the job description, post it in something like an ats, distribute it out into the Internet or some advertising network, wait for candidates to appear, scroll through the list of candidates, score the candidates, try to screen the candidates, talk to some of the candidates, ignore a whole bunch of the candidates, interview some of them, schedule, interview. Right. You know, all of these steps, there's probably 20 of them, are separate and talent acquisition teams have to build this army of people to do all this with different tools for different steps. And the applicant tracking system is sort of sitting behind the scenes here, like the learning management system, but it doesn't really automate all these things. And the experience for a candidate is miserable. They apply and they don't hear anything. They don't know what's going to happen next. There's nobody to call. Maybe there's a recruiter and if there is, then the employer's spending a lot of money on that anyway. Put all that together into a conversational recruiting process and things go 10 times faster. I mean, at McDonald's you can be standing in line to get a hamburger, apply for a job and accept a job before your hamburger is delivered to you. Now, that sounds a little bit ridiculous, but that's the kind of application that you can do with Paradox if you're a high volume recruiting company, which there are a ton of them, retailers, hospitality companies, transportation companies, healthcare companies, all sorts food service companies, all sorts of franchise businesses, have to hire people seasonally, regularly, constantly, because they have high turnover. When I talked to the head of HR at H& M about this, you know, these companies need a way to get people in the door who are qualified to quickly. And the faster they can get them in, the more revenue they can make and the more sustainable they are. So this is not an HR thing, this is a business thing where this system is a business enablement system, just like the sales automation system or the supply chain system or whatever else you consider to be strategic. Now, over the years that I've known Paradox, and I've known them for a while, in very close partnership with them the last three years or four years, they built this out step by step and slowly but surely acquired a thousand customers. A thousand. And these are big companies doing big deals. So I don't know the total revenues of Paradox, but I'm sure it's well north of 100 million a year. This is not a small company and I frankly thought they were going to go public. But you know, so much for Workday. This is a good deal.
[00:05:35] The management team is very savvy. They have great product people, great marketing, great sales force. They go to market in a very unique way. They don't just sell the product and the features. They work with the client to understand their recruiting process and directly prove to the client how much money, time or quality they're going to save by implementing Paradox. There's no question that there's an ROI here before they even do the deal. And that means that Paradox can be implemented fairly quickly because the buyer and the Paradox team have scoped the problem and where the business processes can be improved before they even sign up for the software, as opposed to buying the software and then hiring a consultant and hoping you can get it to do what you thought it was going to do. So many, many, many good things here and the management team is smart, very focused, and I consider to be just great people. So they're going to fit really well into Workday. Second story for Workday is the frontline worker market. We're in the middle of writing a big paper on this, so I've done a lot of research on it and we just gave Workday a bunch of data which they were not aware of, which essentially shows that roughly roughly 70% of workers in the United States work in a frontline capacity. That's 125 million people. These are the nurses, the retail, the restaurant, the Uber drivers, all these people that you interact with as a customer or consumer that are the ones that are generating the revenue, generating the customer experience, generating the services for all of the companies you do business with. Very, very important workforce that has been more or less ignored by HCM providers, with the exception of ukg, who totally gets this. But most HCM providers don't think about this part of the workforce that much. It's an add on with some features, but not really a core. This is where Paradox thrives. So now Workday, in the recruiting domain at least, and probably later and more as I'll talk about, has the potential to win over this segment of the workforce, these kinds of companies, these industries, these organizations in a very significant way. And given the fact that the HM market is very competitive and Workday's growth is, you know, not small, but it's certainly not at the rate it was when they were younger. This opens a lot of new doors for them in customer sets, in industries where they really didn't Play. And you know, in the past, Workday was very apologetic about their industry strategy and pretty much said, we're not going to go after those kinds of industries. Well, now they can to a large degree, and that's a big growth factor for the company that could double or triple the size of the company if they get everything else wrapped around it. So that's number two. Number three is the management team and the technology. Now, you know, Paradox is a little bit pretty chatgpt in terms of under the covers, but. And I've heard complaints from other vendors that it's kind of old technology and this and that. And you know, maybe it is, maybe it's not. I can't really answer that. But Workday needs all the AI skills they can get. Anybody who understands and has experience building AI systems has experience with the user interface, the technology, the design, the implementation process, the use cases for clients. And so this brings all of that expertise into the Workday product group and presumably into the sales organization. Hired Score was a great tool. It's a relatively limited system compared to this. And so you couple this to Hired Score to some of the other things that Workday's done and their talent acquisition system and later the rest of the HCM system can really add this conversational stuff much more significantly. I also think the Workday assistant strategy probably still needs some work. I know they've worked on it a lot, but there's a lot of direction where that can go. If you were to compare that to Juul from SAP, and that's where Paradox could contribute as well. So I think this is a very big deal. I think in terms of market cap and revenue and value, it could be worth a billion dollars to Workday. I don't know what they're paying and you know how that deal is negotiated. But I really am a big fan of this and because I know the Paradox people and we've talked to, you know, dozens of the Paradox customers, they're getting a very, very happy customer set. The only sort of downside to this is the same downside that Oracle had when they bought Taleo and Workday had when they bought Hired Score is that anybody who's using Paradox who isn't a Workday customer could be stranded at some point in the future. I don't think Workday wants to do that. But you know, over time, as these products get more integrated, the non Workday customers will find it a little bit less attractive and they may lose those customers or some of them. But I don't know if that has anything to worry about at the moment. So I'm a big fan of this and hopefully other people see it the same way I do. Second thing I want to talk about is AI transformation in general. Everybody's been dealing with this. There have been, you know, thousands and thousands of articles by everybody on the planet about jobs disappearing, work being redefined, task analysis, the role of humans versus machines, trust and bias, on and on and on. We're going to be working on this for a decade, I believe. But so we started a group of companies to talk with. We're partnering with Rejig here, which is a new company that does automated AI based task analysis for work redesign. By the way, who's a Galileo partner, so you can actually experience rejig in Galileo. And we've had about 40 companies involved. And this week when I was in Europe, I talked to seven or eight companies yesterday about what they're going through. And let me give you sort of my high level thoughts and we're going to write a paper on this. There's sort of some business issues here and then there's some tech issues and design issues. On the business issue, virtually every company I talk to is going through some form of productivity initiative or downsizing or hiring freeze around AI because they see the economy slowing and they want to spend more money on automation and they believe they would be left behind if they didn't do it quickly. And in a lot of companies, the bigger ones, it starts with the Microsoft Copilot or ChatGPT, and expands out and there's a initiative to build AI fluency, or literacy as it's called in the general workforce. And then we start rolling out tools. And a lot of them now are building their own chatbots, not just using the copilot or customizing it, of course, and putting together rules and policies to govern what should go into it and what shouldn't go into it, so that it starts to become a trusted agent inside the company. So one of the banks I met with this week has done a really spectacular job with this. They started with a customer chatbot that was so successful they just rolled it out to the employees. And now they have in HR, three gurus integrating it with workday on their own without the use of a big technology integrator or anybody else. So this is really easy to do, and I've talked about this a lot, is that these are tools that you can do a lot of programming on your own. You don't need to have million dollars of consulting help to do a Lot of this work. Galileo is proof of that. Look at what we've been able to do in a couple of years with Galileo. And we're not even a tech firm, so there's that. The second phase, of course, is people bubbling up or flowers blooming. It's called use cases, where somebody found something really big that's really good for them in their job and they tell everybody about it and they say, hey, let's just put that into a standard application for everybody else. So that requires you to have a governance process so that as new ideas are generated, we can filter through them and decide how much resource to put into which one and double down on the really big bets. I talked about my meeting with Jackie Caney. We're going to have her on the podcast. That's what they did at ServiceNow. That's what some of the big banks I talked to are doing. And that means you need a governance policy or a governance team or a steering committee that can sift through all the things that are going on in different functional areas. By the way, it probably should be done at a business level, not at an individual level. So in hr, there's a bubbling up process in recruiting, in sales, in marketing, et cetera. And then each functional area can come forward with projects for this steering committee to fund with IT support. And eventually what happens, as Moderna has shown, and Moderna seems to be getting an awful lot of press for some good stuff, but they're not that big of a company. So I'm not sure everybody can do exactly what they did is you end up with the IT people building some standard tools so everybody else can build on top of that, as opposed to everybody starting from scratch. And that's what it is good at. So that's sort of happening out there. And then the third level is where a business executive says, okay, well wait a minute, there's one gigantic opportunity here that we need to work on. Let's marshal an even greater amount of resource on that. And maybe that's claims processing and insurance company sales management in a software company, supply chain management in a retail company. And let's build an end to end AI strategy for them. And this is the way it's going to go. It's going to go piece by piece, step by step. And some of you are going to do level one automation, some, which is individual assistance. Some of you are going to do level two, which is automation, individual automation, work automation. Some of you are going to do level three multifunctional agents, and some of you are going to do level four, which is autonomy. While that happens, you're going to be redesigning jobs. Now, there's a fairly inflammatory I consider it article in the Financial Times this week where Laszlo Block is cited as stating that 80% of HR jobs are going to go away. I mean, that's kind of a dream. I don't think that's going to happen, but you will see, and we're already seeing a lot of roles that exist that are no longer needed. If you just think about marketing or publishing or any form of content creation where you have writers, editors, layout people, designers, graphics people, you're not going to need as many of those people, no question about it. So there's going to be some reductions in the number of people needed. And so rather than engineer that from the top, you can engineer it in the functional areas as the implementation of these tools takes place. Because the tools are changing so fast, it's not like you can pick a tool and suddenly decide how you're going to reorganize everybody around it. Because the tools just aren't that mature yet. A lot of them are individual agents and small mini applications. And because these are such easily developed applications, you can build on top of them. Even going back to the Paradox story, every Paradox customer I've talked to has a different implementation of Paradox doing different things because these tools are so flexible. You know, the ERPs or the traditional management tools that we bought in HR in the past, the transactional systems were really not that customizable. That's the reason ServiceNow is so successful, is because it's a workflow oriented system. But the big ones are not really designed that way. Well, this thing is. So whatever AI solution you buy, it may start to do one thing and then do more over time. So we're kind of building platforms for development, not just platforms for automation. Now, one of the sad stories of hr, and I've been sort of warning people about this, but it's turning out to be true, is we tend to do a lot of projects that are architectural in nature. The most significant one that's going on right now is job architecture. Let's create a job architecture, a bunch of titles and levels and functional families. And then we'll know job by job, who's ready for progression, what the skills are of each job, easier to hire, easier to pay, et cetera. I mean, that doesn't really work anymore. It helps, it's needed. You gotta do something like that. But the actual job titles keep changing. Now that we have AI involved I guarantee you you're gonna change a lot of job titles. I mean, the one that I remember the most vividly is when social media manager became a job. And in the early days of that I thought it was insane. You're going to create a job for somebody to post junk on Facebook? Well, yeah, it turned into a career. It's actually a really big job. Wasn't in the beginning, it is now. So there's going to be a lot of that going on. And that's an architectural thing. Skills based organization, famed project that I warned people not to do became an architectural project. Let's build a skills taxonomy. Yesterday I'm in a meeting with seven or eight large companies and we started talking about this and I said, I'm sorry, hopefully many of you didn't spend too much time on your skills taxonomy. And about half of them raised their hands and said, yeah, we did, we're sorry. We started because it didn't really go anywhere or it didn't get us to where we wanted to go. That turned out to be a very big distraction. Useful in some respects for some companies, but not for many because now AI does a lot of that for you and can deterministically quickly change over time without you having to redo the architecture. Well, there is a trend in AI to do task architecture, task design, task analysis. You know, job task analysis is been around a lot longer than I have and it's a, it's a relic of the industrial age. And the reason I think it exists, and I have a big chart on this, I'm going to put this into a big paper. You know, in the industrial age we had to, we had an industrial process that was built on scale, not flexibility. And so we had to, to create jobs in a sense, to do collections of tasks. So what we would do, and I have a visual of this that I think you'll find very interesting, is you had a process of some kind of business process, sales, manufacturing, marketing, whatever. And you'd look at the process on a whiteboard and you'd say, task, task, task, task. You put in a whole bunch of tasks, then you clustered the tasks together into jobs and gave the jobs titles that created a bunch of jobs around these tasks. Well, if the tasks go away because they're automated, what happens to the jobs? They have to change too. I mean, the job of managing the AI didn't exist before, but that job replaced a whole bunch of task oriented jobs that the AI is doing for you. So there's a trend of companies saying, let's do a task analysis of all the jobs in our company so that we can re engineer them. The reason I think that's not a good idea is not that it isn't a good idea in an individual business area. It certainly is, because then you'll really see how this automation works. But across a big company, it's going to change pretty consistently over time. And you can use tools like Rejig, for example, or another tool that does it is drop D R A U P and you can see the repetitive tasks and repetitive jobs around the company and you can get a really good high level perspective on things that should be centralized or automated or tools that should be shared. And so from on the positives of this, you could find areas to automate that you really didn't even need AI for by doing job task analysis of the whole company. The reason I'm not sure it's a good idea to do too aggressively is it may not get you to where you want to go because these AI tools and AI vendors are so creative in the way they're rethinking work that you almost have to start from scratch. I mean, the way I'm seeing it in our company is the tasks that we do today are changing almost every week as we learn how to use Galileo better. So if we did an analysis at one point in time, it'd be very valuable, which we kind of do regularly. But you want to do it in a dynamic way, so you want a dynamic job redesign process, which to me is more of a muscle than just a project. And so I'll talk more about that in the paper. But relative to Workday and Paradox, they're going to be getting into this. And so those of you that are workday customers, I know a lot of you are, we're going to be pretty excited about this. By the way, the flowwise application I talked about last week is probably going to dovetail into this because if Paradox gets well integrated into Workday, there'll be development tools on top of it. So there's a lot of good possible things here that workday could be doing. Okay. So that's a little bit more time than I usually spend on these podcasts, but let's stop there. And we'll be filling you in much more on this at the HR tech conferences coming up and reach out to us if you'd like to talk about your situation. Thanks a lot.
