You Have More Than 80 Different HR Technology Systems? Wow. What To Do?

September 02, 2023 00:22:19
You Have More Than 80 Different HR Technology Systems? Wow. What To Do?
The Josh Bersin Company
You Have More Than 80 Different HR Technology Systems? Wow. What To Do?

Sep 02 2023 | 00:22:19

/

Show Notes

This week I want to talk about the problem of proliferating HR Tech platforms in companies. It's not a simple problem, but it really is everywhere. The average large company has 80+ HR tools and many global companies have twice that. Why is this taking place and what can you do? And how will AI help? Here are some of the answers, and more resources to help below: Resources HCM Platform Excellence: SAP, Oracle, Workday Implementations In Detail The HR Tech Workshop, Certificate Course in The Josh Bersin Academy The HR Consultant Masterclass with Bill Pelster, The Most Popular Course in The Academy The Role Of Generative AI In HR Is Now Becoming Clear Understanding AI in HR: A Deep Dive  
View Full Transcript

Episode Transcript

Speaker 1 00:00:06 Hey everybody. Today I'm gonna tell you the story of a company that has 153 HR technology platforms. And I'm not gonna tell you the name of the company, it's not really relevant, it's a big brand company that you know, but it's a very, very common situation. And the reason this happens is what we call the kitchen drawer problem. If you've looked in your own kitchen drawer, and I'm sure all of you have, every time you pull it open, you look at all the stuff in there and you say, wow, how did this all get here? Why do I have so much stuff? Who's using all these things? How do I get rid of this stuff I don't need? And then you close the drawer and say, ah, maybe some other time. And that is exactly what happens in HR technology. You start to buy more things, there's more innovative solutions, you think they're gonna be more useful to more audiences in the organization, and you proliferate more and more and more platforms and tools and pretty soon it's impossible to shut anything off because you know there's some people using it. Speaker 1 00:01:10 And if you replace it with something else, you're not sure what you're gonna replace it with. And the number of HR platforms goes up and up and up. And you know, the Okta study every year that looks at the number of various different systems that companies use, found that it was around 80. So this particular company had almost twice as many, but I know a lot of you have this. And the person who's responsible for this strategy, the IT person in HR at this company, told me she thinks there's a whole bunch of systems she doesn't even know about. So it's even greater than this. Now there's essentially two fundamental problems that occur when you have all these systems. The first is, of course, on the employee side. It's impossible for people to find all this stuff and figure out how to use it. She actually told me that in her case she had to do a big executive comp study and it took her four hours getting up at 4:00 AM to find the information she needed, get it across the organization just to do the study without even taking the time to do this study. Speaker 1 00:02:09 And employees simply will not use things they can't find. They'll waste their time, they'll give up. And as most of you know, we're now at a stage where if the employees don't see instant value, they simply will not use it unless they have to use it to get paid. So there's this issue of the employee experience. And then on the other side, there's the issue of data. Every one of these wonderfully selected important platforms, stores, other, you know, different data about your workforce, about your employees. And rarely do companies have a way of integrating it together so they can't make good decisions or run analyses or do what if scenarios. Let's suppose there's high turnover in one group. Let's suppose another group is underperforming on their business metrics. Is it because of the management? Is it because of tenure? Is it because of d e i issues? Speaker 1 00:02:58 Is it because of lack of training? Is it 'cause of lack of skills? I mean, those are big questions that are sort of C-level questions that are very, very hard to answer without integrated data. Now, on the data side, you know, I'm excited to tell you, and most of you may know this, that there are some really amazing platforms that can stitch together all of this data using off the shelf E T L tools. Basically extraction tools into a single database with all the metadata defined for you from companies like Visier, who's the leader, one model, who's another sort of fast follower doing really well, another company called Crunch HR out of Europe. And there's a few others, but in the most cases, you know, you're sort of stuck with this distributed data situation. And so the users have the problem of many applications to find and and figure out how to use. Speaker 1 00:03:51 And then you as analysts and business partners and business people don't have access to the data to make good decisions. And I'm not even thinking about things like surveys, teams, slack and all of the data that they create. There's a lot of things that aren't even probably included in that 153 systems. Okay, so what are we gonna do about this? It's HR technology season. There's gonna be a hundred new announcements within the next 60 days from different vendors, and you're gonna suddenly be told to buy all these new AI platforms. Well, I would say there's a series of options here, and let me go through three scenarios. The first of course is for anybody who has a situation like this and the company has decided there is a business reason to fix it. Then there's a process of determining the most important employee journeys and use cases and building a strategy to simplify their implementation. Speaker 1 00:04:46 That may involve buying a tool like ServiceNow or Applaud or First Up or Work Vivo or some of these other employee experience platforms that allow you to build a layer on top of the chaos and leave the chaos in place. You know, when I was at Sybase, we used to joke around all the time that a legacy system is a system that works. And if it really works, why get rid of it unless it's really expensive to maintain or to pay for. I think these days, since we're not really owning any of our software anymore, we're leasing it in a sense. It is expensive to maintain. And you know, if you look at your recurring, you know, credit card bills for all the stuff you're paying for as a company or as an individual, you do have to question yourself whether you're really gonna be using it all and whether you need it all. Speaker 1 00:05:34 So going through a process of thinking about the really important journeys and what's the most difficult will help you to simplify the process. Now, ServiceNow, for example, has made a multi-billion dollar business out of this being a platform of platforms, as they call it. And they're continuing to be very successful at that, in doing a really good job of helping you build portals and other interfaces that sit on top of these legacy systems. So you don't have to, you know, kind of keep replacing them all the time. But ServiceNow is expensive. It costs as much as an E R P. So that layer of software is in itself another legacy system that has to be maintained. So that isn't always the best solution. Second solution that is beginning to pick up speed here is ai. Now, how does AI play into all this? Let me make a couple of comments about what's going on in a, we're gonna AI and I'm gonna talk a lot about it over the next 60 to 90 days at the various conference. Speaker 1 00:06:30 I'm going to, first of all, you know, there are multiple, you know, generations of AI systems. There's what I call bolted on stuff, generative AI that's bolted onto existing applications. There's first generation, there's second generation. The E R P vendors, Oracle SS A p Workday, for the most part are using AI to make their systems easier to use. In other words, rather than you logging into Workday and seeing a blank screen or Oracle SS a p, the system should know who you are and recommend certain things that it thinks you're ready to use and do based on the situation that it knows you're in, the information that it has about you. Now they don't have that much information about you, but they have a reasonable amount of information about you. So those things will make these systems slightly more personalized and slightly more usable. The second of course, more interesting part of AI is the talent intelligence platforms. Speaker 1 00:07:26 Now, talent intelligence, as you've heard me talk about many, many times, is this domain of taking deep, um, neural network analyze data about employees and workers and job candidates and building models, statistical models, AI models that allow you to do lots and lots of things using data. And as, uh, most of you know, these kinds of systems are in some ways the polar opposite of traditional transactional systems in a traditional transactional system. You put data into it and as it starts to fill up with data, it gets a little more intelligent about what's going on with these big talent intelligence systems. They start with billions and billions of profiles and they figure out using the statistics of neural networking, what are the models that are taking place within them to determine skills proficiencies, people who are in similar or adjacent skills, skills that are trending up, skills that are trending down. Speaker 1 00:08:28 Many, many, many other things about people. And these talent intelligence systems have the potential, and I just saw one on Friday that I wanna talk to you guys about to completely change the user experience for HR people and employees in a way that the traditional ERPs probably can't do from here. Now, the reason I say that, not because I don't have trust in the E R P vendors, I think they're very smart people, but their architectures were not designed like this. They're basically workflow and traditional transaction processing systems with AI things added on. Whereas the big talent intelligence platforms are core large language models like chat, C B T, and they can do things and frankly, that the traditional systems can't do. In the case of the user experience where you have lots and lots of systems, the LLMs can do many things. First of all, when you use a system that is a talent intelligence system by design, and I think these systems will become what I call transactional talent intelligence. Speaker 1 00:09:34 In other words, they will start to compete directly with the ATSs, the LMSs, the H RMSs. They can do things that are much, much easier for the user. Let's suppose for example, you're trying to do a succession plan for a job or a whole bunch of jobs and you, you have a role that's been described, you can, you know, get the system to describe what the role is. The system will determine the skills of that role from the role description, also from the people that are in the role. So it does a very intelligence job of figuring out what this job is all about in a way that your mind probably couldn't do very easily. And then it will look through the entire database of people in your company and it will give you a list of this people that are most likely to be immediate successors or adjacent successors. Speaker 1 00:10:24 And for those that are a couple of steps away, it will give you a development plan of the skills and capabilities and experiences that they need. And then it will point you to the content in your content library that possibly would fill those gaps. Now, in a traditional transactional system that's takes a lot of work on your part to do that. They're not designed to do that much work for you. They're more designed as kind of reporting tools. So what I call these transactional talent intelligence systems are going to be much, much more functional, much easier to use, and you won't have to go out and buy as many small platforms. By the way, you know, one of the reasons that companies have 153 systems is because the market goes through phases when there are new ideas that come in the HR domain, lots of vendors build standalone tools. Speaker 1 00:11:18 I remember very well 10, 15 years ago there were 25 vendors selling performance management tools, 25 vendors selling learning experience and learning management systems. A bunch of vendors selling survey systems. Again, probably 20 to 30, a whole bunch of companies started to build analytics tools. Well, that stuff kind of collapsed into what we at the time called integrated talent management. Most of those vendors got acquired or merged or got backed by PE firms and then bought each other. And we ended up with this kind of talent management systems market. Well now we're going through the same thing. The generative AI tools, the advanced tools for interview intelligence, the advanced tools for sourcing, the advanced tools for skills assessment, for skills taxonomy and so forth are are sort of standalone. And you can't buy them all from one vendor yet. I mean, the big vendors kind of are trying to build all that stuff, but they can't, they can't do it fast enough. Speaker 1 00:12:15 So we're, we're kind of in this new world of the proliferation of vendors in all of the new domains. And I think these, what I call transactional talent intelligence platforms will become major, major systems. And I'm talking about companies like Eightfold, gloat, Beamery, phenom perhaps, and some others that haven't really grown very big yet that are gonna become much, much more interesting than you even think they are today. The third thing that has to do with this business of proliferating systems is the front end chatbot. Now we are doing a huge amount of r and d here in our own little company on this, and it's really very interesting what we're doing because we're basically a content company. So we in some ways have the same problem you do in that we have 800 research reports tagged by topic and different industry and so forth. Speaker 1 00:13:08 And the poor users gotta find all this based on whatever their question is. And you've built, I don't know, I'm, I think I've built six or seven websites in the last 25 years and none of 'em, none of 'em are ever perfect. So whereas the chat interface is actually a pretty interesting solution. I talked about this in the podcast a week or two ago, but what we're now finding in companies and in the research we're doing for our own case is that what you can do with generative ai, by the way, there's a large article on general of AI I just published this week that you can read through on all these use cases is you can take a transactional system like Workday, Oracle or S A p, and you map the user transactions or user experiences into the chatbot and let the user find the system or interaction they need through a chat interface, through asking questions without having to poke around and get into the core systems. Speaker 1 00:14:04 I B M has done this very well. SS A P's doing this big project with both Watson and OpenAI to do this for SuccessFactors and the rest of the SS a P systems. Workday is starting to work on this. I think we'll hear more about them in the next couple of weeks, but your environment is bigger than that. You've got whatever the core system is, and then you have the a hundred, the other 154 things. So that's gonna be a big solution. By the way, since most of this isn't available off the shelf, and most of you are not AI chatbot developers, what's been going on is a series of what we call hackathons or Prompt aons. Three or four companies we've talked to in the last week are now doing design projects inside their company, taking HR people and setting them loose to come up with the use cases for these employee experience AI-based applications. Speaker 1 00:14:56 Because this isn't a problem, as I talked about in the article I published yesterday of just turning on open AI chatbot and assuming it's gonna answer all these questions, you really need to make a lot of decisions about who are the users, what data are you gonna put in there, what systems are you gonna map into the orchestration part of the system and what problem you're trying to solve. I mean, maybe someday there will be this Uber chat bot that does everything you ever wanted to do in your company, but we're nowhere near that yet. It's too complex under the covers. So before you could scratch together this idea and start trying things, I think it's well worth deciding amongst your teammates, where do you think you're gonna get the greatest bang for your buck? And then start evaluating and looking at tools and solutions to build a conversational interface. Speaker 1 00:15:45 We can show you what we're doing and we can explain how this works. It's not super complicated, but there's lots of moving parts. And I would add that these prompt aons are very useful. I'm, I'm gonna write an article next week or so on this and give you some details on one of the ones that's really interesting. They allow your HR people to learn about AI and think about it and contextualize how they believe it could be used in your company to make work life better or to help people make better decisions. And there are gonna be a lot of opportunities. I know enough about AI to be very convinced that sometime from now maybe 2, 3, 4 years from now, all of these data systems we have that will continue to have disconnected sources of data on different parts of the business will be integrated into a massive L L M. Speaker 1 00:16:38 And you'll be able to ask it questions like, who are the business units or geographies that high have the highest turnover this month? What are the number one, number two, number three, correlated factors in the talent environment that seem to be correlated with this turnover, et cetera, et cetera, et cetera. And those are queries or questions that are going to be worth millions and millions and millions of dollars to your company. Because if you find the answer to any one of those questions that's reasonably correct, the solution that you implemented have a huge R O I and I don't think we're gonna be too far away from that. I mean, this is gonna take some work on your part working with it. So that's kind of the state of the market. The final thing I would say about this issue of having a lot of systems is not only is it a cyclical problem that comes and goes as the market changes, it also has to do with the economy, by the way. Speaker 1 00:17:36 When the economy really takes a nose dive, nobody buys anything and people start shutting stuff off. But it's also just the nature of the beast. The complexity of HR is maybe one of the most complex areas of business I've ever worked in. All of the issues of recruiting, all of the issues of training, all of the issues of employee experience and surveying and employee communications, diversity and pay equity, identifying who's capable of what role, moving people into new positions, uh, engaging them along the way. Those are very different complex problems that vendors have spent a lot of time on, including by the way, assessing employees, the, the assessment market and so forth. Just talk to a big company about that and you're not gonna find one solution that can solve all those things. You never are. So this is a constant process of looking at the technologies you have, architecturally deciding which ones are enterprise class, which ones are not, which ones you're going to collapse into other vendors, which ones you're going to let stay and which ones you're going to delegate to a local organization to take care of on their own and pay for on their own. Speaker 1 00:18:47 There's a lot of situations where a country or a business unit or a team wants something and they're willing to pay for it and as long as the IT department is comfortable with whatever integration it needs, you just be done with it and it's their problem and you may not be able to shut it down. Now, I know big global companies like Microsoft and Nestle and many of the S A P customers rely on one core platform like SS A P and they really don't let people go around it for a lot of these add-on tools. And so if they buy an add-on tool, they make sure that it is well integrated with that core platform. That makes a lot of sense, but it's very hard to do because the core platform's, APIs and their business partnerships are limited. So you can't accommodate a high volume interaction with every single platform in the market that you find. Speaker 1 00:19:42 So this is a lot of ongoing work. We have a couple of resources to help you here. Two things I wanna point out. First of all, in the org design class, the super class that's been going on the last couple of weeks, you'll go through a process of getting to know a case study company and diagnosing the stakeholders in that organization to figure out their business needs. And it's a very, very well designed simulation that will give you a lot of real time experience in how to do stakeholder analysis and build a great business strategy for any new implementation of any technology. So that's one resource. The second thing we have in the J B A is a course called the HR Technology Workshop. And what it is is it's a four hour course that will take you through the fundamentals of all the different aspects of HR technology and you will actually go through a workshop exercise to think through how you start prioritizing the uses of the various systems in your company. Speaker 1 00:20:46 But basically that's not gonna solve the problem for you. The ultimate problem is you working these things out with your, with your team and with your IT department. We also do a lot of this work for clients and we're happy to talk to you directly about your situation. And I promise you, I'm gonna give you a lot of new ideas and concepts on this in these two speeches I'm giving in the last couple of, uh, of October at the HR Technology Conference. It's a really, really disruptive and, uh, good time to take a kind of a fresh look at your HR tech stack and consider all the new tools that have come out from Microsoft, all the tools that are coming out from the core E R P vendors and then all of these wonderful new AI startups. Take a look at the article I just wrote on generative AI this week that will also help and I look forward to hearing from you. 153 systems is a lot, but I wouldn't be surprised if you work for a big company that you're not too far away from that number yourself. Thanks, you guys talk to you next week.

Other Episodes

Episode 0

February 17, 2024 00:32:02
Episode Cover

Interview With Joel Hellermark, CEO of Sana - AI-Powered Learning Arrives

In this podcast I interview Joel Hellermark, the founder and CEO of Sana, one of the most exciting new learning and AI platforms in...

Listen

Episode 0

August 10, 2022 00:20:54
Episode Cover

A New Role For Corporate Learning? Yes: Growth In The Flow Of Work

In this podcast, I discuss our groundbreaking new research on corporate training, Growth in the Flow of Work. After almost a year of research...

Listen

Episode

May 03, 2024 00:27:11
Episode Cover

Rethinking Employee Experience: Lessons from The BBC and Schiphol Airport

This conversation explores the concept of employee experience and its evolution from employee engagement. The case studies of the BBC and Schiphol Airport demonstrate...

Listen