Episode 186: How IBM Uses AI to Transform Their HR Strategies (Interview with Nickle LaMoreaux)
In this episode of the Digital HR Leaders Podcast, David Green is joined by Nickle LaMoreaux, Chief Human Resources Officer at IBM, to explore the transformative impact of Artificial Intelligence (AI) in HR.
With a spotlight on IBM’s pioneering work in integrating AI into HR practices, Nickle shares her invaluable insights into how AI is revolutionising the way we think about talent management, employee engagement, and predictive analytics.
Key topics discussed in this conversation include:
The evolution of AI in HR at IBM;
Examples of how IBM is harnessing AI to transform HR practices;
How IBM turned the challenges of implementing AI in HR into opportunities;
How IBM ensures ethical and responsible use of AI;
The emerging trends and the skills HR professionals need to stay ahead in the digital age.
Support from this podcast comes from Worklytics, a people centric analytics solution that combines passive listening with Organisational Network Analysis (ONA) to help you understand how work is getting done.
Curious to see how it works? Worklytics is offering a free Collaboration Analysis to the first 10 qualified companies who express interest by clicking on the following link: www.worklytics.co/DigitalHRLeaders
[0:00:00] David Green: Four dynamics impacting many HR functions in a significant way today are, first, cost pressure. HR is being asked to do more with less, which means that there is pressure to optimise HR operations. Secondly, many important HR processes have to be accurate: payroll, benefits, and onboarding to name but three. Thirdly, employees want consumer-grade experiences. They want the same level of personalisation and experience they enjoy as customers in the workplace too. And fourth, executives increasingly expect HR to be a deeper, more insightful partner throughout the business.
As Nickle LaMoreaux, Chief Human Resources Officer at IBM, and my guest on today's episode, explains, all four of these factors were converging on the HR function at IBM. IBM is deservedly regarded as a pioneer in developing and implementing cutting-edge technologies. And under Nickle's leadership, their HR function has been a shining example of how AI can be harnessed to revolutionise everything from talent acquisition and learning to employee engagement and workplace culture. In our conversation, Nickle will share insights into the strategic integration of AI within HR, the impact on IBM's workforce, and how they are navigating the challenges and opportunities AI presents. We'll also explore practical examples of AI-driven initiatives at IBM, delve into the ethics of AI in HR and responsible AI, and discuss what HR professionals need to know to stay ahead in the rapidly evolving age of digitisation. With that, let's get the conversation started with a brief introduction from Nickle herself.
[0:02:05] Nickle LaMoreaux: I started my time with IBM actually as an intern. I would say I stumbled into the Human Resources profession. I was very intent on becoming a lawyer and I wanted to go to law school, but I didn't have any money. And so, for me, the HR job that I got at IBM was going to just be a slight detour while I earned a little bit of money. I know I could have picked a different profession to maybe earn a little more money, but I picked HR and I started working with IBM, and I will tell you that I fell in love with the profession from the business impact that you could have, the people impact you could have, and then doing it at a company like IBM, which is hardware, software, services, 250,000 employees in 170 countries. Before I knew it, I was a couple decades in and continuing to love the domain.
I did study under Diane Gherson, my predecessor, and learned a lot about analytics and AI from her, which I know we'll talk about a little later here in the session. But across my career, I've been in virtually every HR domain. I did have the opportunity to do two international assignments, both in China, during my career, and I took over in September of 2020 into the CHRO chair. So, a little bit of an auspicious time to be taking over the role, but quick study and quick learning; that's what I would say about the last three-and-a-half years.
[0:03:45] David Green: And it's interesting because I think a few of us stumble into the HR profession, and that for many of us, once we're in there we don't leave it. And actually, one of the things I think we'll talk about throughout our conversation, Nickle, is that business focus that you have in the HR function at IBM, which maybe not all companies have. Maybe some companies, the HR team is a little bit too much about HR, but I know that's very different at IBM and we'll talk about that. So, over the years, obviously you've been in the role now for four years and obviously worked closely for others that were doing the role beforehand. How have you seen the role of the CHRO change, particularly now that we've entered the age of AI and particularly now that I think in more companies, HR has become a strategic business partner maybe than the support function it was years ago?
[0:04:36] Nickle LaMoreaux: Yeah, so I know, David, maybe believers like you and me and others have always said that for most companies, product is not their strategic differentiator, it's their talent. They can still have a great product, an above-average product, and I'm not saying that's not important to our business model, but really what is a differentiator, because you can, over time, start to copy business models, business processes, product? Over time, in the long term, what really is a differentiator is your talent. And I think there were some believers early on, but I think in this age of disruption, in the digital era, where business models are starting to get on an even playing field, more and more companies are waking up to this idea of, it is about the talent, it is about the people. So that is a big shift that I think is happening.
I think the second thing that has happened is, I know many CHROs and HR functions were thrown into it, but we really earned a lot of credibility during the pandemic. If you think about health and safety of our employees, if I think adapting to new ways of working in remote work, adapting to differences in worldwide standards around vaccination and travel and shutdowns, and working side by side with business leaders about then how their business model was impacted, there was also this piece that really business leaders started to see us in a new light in our ability to be agile, jump in, use data to make decisions, deal with nuance and complexity. So, those things are converging, which I think are putting CHROs and HR functions in a very different position in virtually every organisation.
[0:06:40] David Green: It's interesting, I think right early in the pandemic, there was an article in the Economist that was much quoted by people, that talked about the role of the CHRO in that crisis being as important as the CFO in the Financial Crisis. And actually, it was quite prescient because it was actually in the early stages of the pandemic and actually came to be the case in many organisations. And as you said, it opened the eyes of executives. So, now that they actually maybe saw what they hadn't seen before, I know that's not the case in IBM, but saw that they hadn't seen before, that HR was actually a really strategic player at the top table.
[0:07:16] Nickle LaMoreaux: Yeah, absolutely. And now, if you kind of fast-forward to, okay, so here we are in AI, the era of AI, massive shift, CHROs are being asked to do a lot, HR functions are being asked to do a lot. They're being asked to transform their own functions, right, more delightful people experiences, consumer grade employee experiences. They're being asked to make sure that we're optimising budgets. Are we using the money that we have in the HR function and that we're deploying, and the investments we're making in talent, are we using it the right way? And AI is helping us do that. They're also being asked to manage this increasing world of complexity and compliance that can vary by country, state or province, city, and it doesn't mean just throw more people at the problem or the compliance work, but how do you use AI?
So, one big piece of this is, CHROs are being asked to use AI to transform their function and the people experiences in their organisations. They're also being asked to reskill the workforce because AI is changing every job. So, regardless of the jobs in your organisation, HR functions are being asked to build retraining and skilling plans so that these professions, other professions, can then meet the needs. And then finally, CHROs and HR functions are being asked to do what I will call almost time-study analysis about what time is being freed up; what does that mean for workforce planning; do you backfill? And that is a really complex piece of workforce planning that is squarely falling on the shoulders of the HR function. So, in this era of AI, I would say we're now again, we're not hand-in-hand talking about shutdowns and vaccinations, we're now sitting side-by-side with business leaders talking about these elements.
[0:09:19] David Green: Yeah and that's something that is a conversation that's going to take several years and keep evolving, because the technology is evolving so fast, isn't it?
[0:09:27] Nickle LaMoreaux: Exactly, we are not at end state. So, what we're re-skilling for is going to constantly be a moving target, where the impact on workforce planning is going to be a moving target as the technology evolves.
[0:09:41] David Green: Let's talk about first a bit about how you're incorporating AI and machine learning into HR at IBM. I know from my time at IBM, even back before the pandemic, that IBM has been at the forefront of this, not just in helping the wider organisation transform, but also in actually infusing AI and machine learning into its work. Can you share, or maybe a sneak peek of some of the initiatives that you've been working on related to AI in HR at IBM?
[0:11:19] Nickle LaMoreaux: Absolutely. So, within our talent, our people experiences, we're really using AI in what I'll kind of call three broad categories. The first, and David, you will remember this, we've been at this a long time, is around recommendations. So, I'll use a very basic concept of recommendations. But this is where we've got AI-driven skills roadmaps or recommendations in our learning platform. When I log in and you log in, we'll see two totally different sets of recommendations because AI knows about my profile and my aspirations versus your profile and your aspirations. We use it in compensation, so when managers open their annual salary budget, that AI has a recommended and ideal way to deploy that budget. Now, managers are still the decision-makers, but it is there to make recommendations and help surface insights. So, that's one area. And we've been at this actually from the very early stages of predictive analytics and neural networks, all the way now through to generative AI, so we've been upgrading that capability. So, recommendations is one.
A second area that we have been using AI is in the area of what I'll call assistants. Some people call these chat bots, as an example, but this is where employees can go in and ask questions and the assistant is surfacing information. What's the vacation policy? How do I transfer one employee to another manager? So, that is where assistants are helping to surface information in very real ways, and we've been at that since 2017.
The last category where we are using AI is what we call agents. So, this is no longer about AI just surfacing information or making recommendations, but where AI is actually doing work. Some of us started experimenting with this and Robotic Process Automation, RPA, but now with generative AI on the table, this is now intelligent automation, where you literally have digital labour, digital workers working side-by-side with HR professionals. So, these are the three broad categories. We can go deep on each one of them if you would like, but that's how we're thinking about transforming and using this technology to transform our HR function.
[0:13:56] David Green: Well, let's do that. I mean particularly, let's maybe start with the latter. I read something recently as a case study I think of HiRo, and I think that falls into the third category, doesn't it, the aging category, and I'll let you tell the story because you'll tell it a lot better than me. But I was super-super-impressed because you could see the benefit for the business, but also the benefit for the HR business partner and the benefit for the employee. So, now that I've set it up, hopefully, Nickle, I'll let you tell the story.
[0:14:27] Nickle LaMoreaux: Yes, so I will tell you a little bit about HiRo. So, if you asked the average HR partner or HR professional at IBM, "Name the one talent process you hate the most", we would hear, almost unanimously, we would hear them saying, "The promotion process in consulting". So, we have a large consulting business that runs an annual promotion process. And this is a process where people get promoted based on set criteria, their billable rates change. Those of you that may be listening that are in some form of professional services, this would not surprise you. And why our HR professionals disliked this process is, yes, it was a volume, we run this process in 170 countries, over 100,000 employees in some way are involved in this process.
But HR professionals would tell us, one, they felt like all they were what we call spreadsheet jockeys, during this process. They were pulling data from our learning systems, our performance management systems, our client feedback system, our skills taxonomy systems. They were pulling all of this data together, managing spreadsheets, sending it to managers who were making notations about who and shouldn't get promoted. Then there were comp recommendations being floated around. Sometimes they were correct, sometimes there was mistakes, payroll dislikes that they had to do rework, compensation was frustrated that we weren't always getting accurate information. It was a long process and nobody involved in it really felt that it was a delightful process. And then at the end of the day, we were getting feedback from managers that the administration of this process took so long that they weren't able to appropriately focus on how do you celebrate somebody's promotion? How do you prepare for a conversation to tell somebody they're not getting promoted?
So, we had redesigned processes, we had tried to make things simpler, we had automated spreadsheets, we had done training and nothing was changing. And so, we felt like we needed a step change. And that's where we looked at the IBM technology, and there is a technology called Watson Orchestrate, which is intelligent automation. It's AI-driven automation, and it essentially is a digital worker. So, an HR professional will decide what the business leader is this year, here are the criteria. HiRo is our digital worker. HiRo will read that document, ingest that document. HiRo will then go out to our HR systems, as disparate as they may be, pull together all of the information automatically, then send lists to 10,000 different managers, customised in email saying, "Nickle, promotion cycle's kicking off, here's the criteria, here's who's eligible in your organisation". You, David, would get a different email. This is all happening without human intervention. The manager can then go back and say, "I'd like to nominate these two people, this person's performance isn't where I'd expect it to be". HiRo ingests all of that. HiRo then, based on criteria we've given it, will make compensation recommendations that the managers can accept or adjust, and then send it directly to payroll.
So then, what is the HR professional doing in this time? They're handling any escalations or complex cases. They are also working with the managers to coach them on exactly what we just talked about, how are you going to celebrate David's promotion; how are you going to have a real conversation with Nickle about what her gaps are and why she's not getting promoted this cycle? And so all of this, it actually saved our managers over 50,000 hours last year, this promotion cycle. And we're getting better results, better feedback from employees and managers, we're getting zero defects on the way to payroll and compensation payments, and our HR professionals are now no longer dreading the promotion process. So, that's one example of where a digital labour, an agent like this, is making a huge difference for us.
[0:19:08] David Green: And I think what's good, I'm going to come back to the HiRo bit in a minute actually, but so two questions on that. Why HiRo? And then secondly, I think what it really shows us is that this isn't about replacing, this is about augmentation. It's about pulling away all the administratively heavy, frankly quite cumbersome and quite boring tasks away and allowing HR professionals to focus on what they do best, as you said, for the escalation examples or the support for the manager, so they can land the good news or the bad news well, and that has a better impact. That's better for the HR professional, it's a better impact for the people manager and the employee, as you talked about.
[0:19:50] Nickle LaMoreaux: Yeah, absolutely, and that's what I think is really important here. At no time is the AI running wild. The AI is not coming up with the recommended criteria for the promotion cycle, they're not blindly making compensation recommendations, right? It's all based on guidance that the experts are given. But what it's doing is pulling those reports from the different systems, putting them together, cutting and pasting and sending different emails to different managers, compiling the file for payroll. That's stuff that HR professionals don't want to do. And so, we always have a human in the loop, and that's really important for us on these agents, and we can talk more about that.
So now, why HiRo? So, two reasons why HiRo. So, the first one is, as we started this process, we thought, "Well, this agent will become the HiRo to the HR functions if we actually land this correctly". But we decided that in HR, for each team that puts in a digital agent somewhere into their process, that we always want to have an H and an R in the name. So, HiRo. We've got Charlie and Sherlock and Hermione and Harry. So, we are really kind of working through any names with H and R, but this is one example. Sherlock may not surprise you. Sherlock is our travel and expense. So, as reconciliations are coming in or there might be a deviation from policy, we have hundreds of thousands of travellers a year, and so how does that work through the process?
Hermione posts job requisitions for us, so builds job descriptions and then helps a talent acquisition professional post those job requisitions. And this is what we're working through of, how can we take, exactly as you said, the administrative burden that exists in all of our jobs, my job, your job, how can we start to take that work away so that we're freed up to do the stuff that we do best?
[0:22:01] David Green: And it's the same, isn't it, with the recommendations? So, if we go back to the learning recommendations as well, I mean as an employee, as an IBMer, you share where you want to get to in your career at IBM or your career aspirations, but it also looks at your skills, doesn't it? It also looks at your skills and maybe what adjacent skills that you've got, and maybe will share information that the employee doesn't know perhaps about what careers and roles are going to be in demand even more at IBM in years to come, and how that can give them a really good career path within the organisation as well. So, it's that balance, isn't it? But ultimately, it's up to the employee what paths they go down.
[0:22:40] Nickle LaMoreaux: Absolutely, and this is another good example where I think that we're trying to think about, where do you infuse AI where it's going to have the biggest impact on the business? And we were hitting up against a few things when we started down this path, especially on the AI skills, imprints, recommendations. So, we had a skills taxonomy assessment that many organisations had, in a very traditional way where employees and managers would fill out the taxonomy assessment on an annual basis. At one point, we moved to a biannual basis. And one of the frustrations that came forward is, we heard from both managers and employees, as soon as we fill it out, a week later it's out of date. Our employees are continuously learning new things or gaining different experiences or being deployed on new projects, working with different clients. And so, how do you keep this updated, because it feels a little bit like a waste of time, so how do you keep it updated in a living way?
So at first, what we did was we used it for what we call skills inference. It was to take everything that it knew about my profile, your profile, as soon as I was deployed on a different client, or got a project assessment back, or it would be looking at papers I published in our research division. And what the inference would do is, anytime it thought that you had built a new skill, you might get an email that says, "Nickle, have you gained a skill in Python? We noticed you took this class, you worked for this client". And I might say, "No, I was just testing out our learning system. I didn't really actually complete it, so, no", or I might say yes. If I then change the skill, it will go to my manager for verification. But it was happening in a much more bite-sized, small way. Employees validating it first in case it had picked up something wrong and then manager signing off. Huge.
But then we went a second step, because then we started to get feedback to say, "Okay, so now the system is smart, it knows about me". But you still have the very standard, very tired learning roadmaps. So, even though David and I are two totally different people with different sets of experience, if we go into a system and we say, "Tomorrow, we want to become a blockchain developer", it's going to give us the exact same roadmap. And people were getting frustrated. I have to take things that I already know or already learn. And so now, AI will generate customised roadmaps. It will take my profile, bump it up against the profile I want to become, and it might say for David, it will take you 10 hours to become a blockchain developer, and Nickle, it might take you 10,000, and here are the different things that you need to do to get there. And those real-time learning roadmaps that were customised to the individual, it does a couple things for you. You get to reskill people faster because it's really giving you just what you need to know; it's creating better engagement with the employees who want to take the learning because they're not repeating something or something that's already foundational to them; it's allowing to get us talent to the right place at the right time. So, that's an example of where for us, this has unlocked a ton of business value, particularly in a professional services business.
[0:26:29] David Green: Did you know that managers often have more influence on our mental wellbeing than our partners? A third of managers feel unable to support their team through mental challenges. OpenUp gives managers the tools to create a safe working environment, so they can support their teams through the challenges of everyday life at home and at work. Wondering what OpenUp can do for your managers? Go to https://openup.com/business/
For listeners, how do you ensure at IBM that the firm uses AI and machine learning tools responsibly and ethically?
[0:27:21] Nickle LaMoreaux: Yeah, so we have done this at a couple of layers, right? So, the first layer we've done it is for all of our AI, regardless of where it's deployed, internally with clients and products that we build, we adhere to kind of five AI principles: explainability, the AI must never be a black box, it must always be something that you explain how it's working, when it's working, where it's working to the user; fairness, so again, we believe that actually AI can help humans, it can assist humans in making unbiased decisions, but fairness is a very key point of the AI. The third thing is robustness. It must be secure. If it's ingesting data, particularly HR SPI data, we need to make sure the platform, the algorithms themselves are secure. Transparency, what data is it using; why is it using that data; where in the recommendations? And then last, privacy, is there certain data that the AI algorithms will not use? So, those are five principles that we are very clear are principles that every AI solution must uphold.
The second thing is then, you must, for whatever the use case is, apply certain principles. So, as an example, in HR, we believe that talent decisions should be made by humans. That's why we've put in the principle, AI will never be a decisionmaker. So, we're very clear about for your use case, for your function, whatever it may be, are you very, very clear about what those principles are that align to maybe your company or your business model? And then finally, and I know many companies have started this, but I encourage all companies, if you haven't done this yet, to consider it, we have an AI ethics board. So, anytime we are going to use AI in the company, it doesn't matter if it's in HR or finance or sales, before that AI can be deployed, it has to pass through this AI ethics board. It's a cross-functional board made up of a lot of different experts. It could be CIO, HR has a representative there, legal. We also have people from the business. And that is the check and balance to make sure that you're meeting these principles and priorities that exist in the organisation, but it is also the test case and the board you need to come back to, to show results after three months, six months, nine months to have a continued licence to use the AI in that way. This helps ensure that what's initially approved doesn't start to morph into something different. So, that's the model we run to help govern this.
[0:30:34] David Green: I mean, governance is such a huge part of this, isn't it? One of the principles that you mentioned around transparency, that's important for all AI, but I think it's particularly important when we're using it in the organisation for employees, isn't it? So, employees know what the AI is, why we're doing it, what data we're collecting, all the other elements that you mentioned around robustness and privacy, but also why we're doing it, why we're doing it from the company perspective, but why we're doing it for employees as well, what benefits is it going to give to employees? Because if employees can see the benefit, they're much happier for their data to be collected and provide their data. Yeah, really, really interesting.
Let's bring it back again to the HR professional now, Nickle. For years, we've been talking about what HR professionals need to learn from a skills perspective. We did some research actually at Insight222 last year actually as it related to people analytics. I think it could be applied to AI as well. AI is particularly, in companies like IBM, it's just part of what we do know in HR and the wider business. What are the new skills that you foresee that HR partners, HR professionals need to develop to stay relevant and add value?
[0:31:53] Nickle LaMoreaux: Yeah, so some of them are going to be our old standbys, and in fact they may get even more important. But the one thing I will say is, all HR professionals, regardless of what domain you're in, if you're a generalist or in any of the functional domains, you do have to learn the basics about AI. I believe very strongly this is going to transform all of our jobs, and so we can talk more about that. But it means then that this isn't going to be necessarily a tops-down, big-bang, digital transformation. It's more grounds-up, how can I use AI to help my process, my programme, my policy; what digital agent would help free up me from these this administrative tasks? So, I think this idea of having a general understanding of it is really, really important.
I think, as a result then, what happens is, particularly if we get freed up from these administrative tasks, some of these higher-order skills, analytics, business acumen, really making sure that you understand your domain, your HR domain, is going to become more and more important. One good example that we use is with our digital assistant, AskHR. We were finding that our HR business partners were spending as much as 50% of their time answering transactional questions, how do I put an employee on a leave of absence; where's the link to the salary plan tool; very transactional questions. If an AI assistant is now answering those questions, that means that HR professionals need to spend more time coaching, having role-playing conversations, doing workforce planning. So, if anything in this, I think some of the core skills we've been talking about in HR become more important, and certainly your domain expertise in your functions become more and more important.
[0:34:05] David Green: How do HR professionals gain that business acumen?
[0:34:09] Nickle LaMoreaux: So, this is one of the questions I actually get asked probably most often from HR professionals, both inside and outside of IBM is, how do you learn this? And look, I don't think there's a one-size-fits-all answer here. I know of some HR professionals that take the approach that they make sure that they listen to earnings calls, that I know others that have taken some approach of let's do learning around accounting and finance. One HR professional said to me, and this is advice I have often taken is, never eat lunch alone. So, they find different people, non-HR people, in their organisation, and they invite them to have lunch with them in the cafeteria to talk about, what's your job as a developer; what do you really do; what's the best thing about your job, the hardest thing about your job? Pick their brains on how they think the job is evolving. That's another great way to do this. And then honestly, for some HR professionals, how they've gained the business acumen is actually rotating into a business job and then coming back to HR.
So, I think that as long as you understand you need it, there are multiple pathways out there to get it. I don't think one is better than the other, and you can do what fits maybe your career aspirations or your organisation model that you sit in.
[0:35:39] David Green: The other thing we found when we did the research last year was, for companies that are trying to build that data literacy, that data-driven, maybe data-informed culture in HR, one of the other things that we found that was really important in companies that were doing that was that there was role-modelling by the CHRO and the HR leadership team. Now, I know that's the case in IBM because I was fortunate enough to work there, but what are some of the role-modelling things that you personally do as a CHRO around using data and analytics in your day-to-day conversations with your colleagues, but also with the business?
[0:36:14] Nickle LaMoreaux: Yeah, and my team will back this up, I am one of the power users of our analytics dashboard. I'm probably in there running reports, manipulating data more than just about anybody else, but it's one good way. I also, if I happen to be on a meeting where we're talking about data, we're not sure, quickly pull it up, show others in the meeting how to do it. So, I think that don't be afraid to do that, whether you sit in the CHRO seat or whether you're an HR executive and HR manager, that role-modelling is really, really important.
I think another thing is, you think about data analytics, and David, I know you know this, there's a plethora of data out there, and sometimes the data can tell you what you wanted to say, what you needed to say. So, I think another good role-modelling thing is always bring multiple people into your data, "This is how I'm interpreting this data", and somebody else may say, "Hey, I think you're missing a key point of this". So, have those active conversations and don't just view the data as black and white through your lens or your interpretation. And then maybe the last thing, and this is more strategic, but I also think prioritisation is important. Sometimes people will say, "Well, I wish we had this data", and sometimes you have to ask the question, "But would we make a different decision? If you would make a different decision, go get the data. If you wouldn't, don't waste time. Don't hide behind the data". And so I think that's another thing that we kind of have to role-model a little bit is that element as well.
[0:38:05] David Green: Where do you see AI going next; what should we be looking out for? And then maybe, where do you see it going next in HR and the role of HR professionals?
[0:39:07] Nickle LaMoreaux: Yeah, okay, so this is a big question.
[0:39:09] David Green: It is a big question!
[0:39:10] Nickle LaMoreaux: Maybe a little bit of an unfair question, David, as we think about, I don't know if anybody knows exactly where this technology is going to land. So, I will just start by saying this. But I think that is really important for us to all acknowledge as we think about the role of HR. So, as we think about the role of HR, because we don't know where this technology is going to land, we have to create a culture of experimentation, right? This is not a digital technology that you're going to big-bang roll out and it's going to be here for five or ten years. So, we have to make sure that in our organisations that we're thinking about, how do we create this experimentation that some of these pilots we run with AI may fail? Or we may decide that even if they're successful, they're not giving us the business value we want, so we're going to stop them. So, that idea of experimentation is really important for wherever this technology lands from an AI perspective.
I think the second thing that we need to focus on as we're thinking about AI in the organisation is going back to the point you made, can we use data to inform where AI is going to have the biggest impact? When people start asking me about, "Well, where do you infuse AI in the organisation?" Your highest-volume areas is one piece that I always give as a suggestion. Your process where it has the lowest satisfaction, right, so that's another area that you might want to start with. So, we think about those things of using data to help determine where you might have some of the biggest impact with AI. And again, I think that's really important for HR to play a role in that.
I think that over time, organisations will settle down. There will be a little less experimentation, that it will become standard in organisations of where you're using AI and where you're not. But it's not where we are right now. And so I think the last role that HR can play is as you're skilling employees, think about this point of making sure that you're emphasising continuous learning. Because you don't want to skill employees and say, "Okay, we're going to end at agents that look like HiRo, so let's just make sure that we're training you on how to use an agent like HiRo". Well, the digital agent, digital labour technology may look very different in six months. And so, you've got to get people prepared of, here's what they need to do for now, and there will be another leg of this journey.
[0:42:06] David Green: That's really good, and actually to your point about experimentation, I think it's really important maybe, as HR's not being a function that has experimented. It's been that we wait, we do these big rollouts and we do big bangs and we try and do it across the whole organisation. But actually with HiRo, you initially piloted that, didn't you, just in the North American, I think, part of IBM Consulting. You saw the results, you saw that it could have a big impact, so you rolled it out across other parts of the organisation.
[0:42:35] Nickle LaMoreaux: Yeah, and I think particularly as you think about experimentation within HR, there are two challenges most organisations come up against. One is this idea of, anything we've done in HR with technology to this point, think about HRIS systems, talent acquisition systems, even engagement platforms, they have been big bang. Your team works behind the scenes on it, then you launch it to the organisation, after a year, two, sometimes three years of working on it, it's a multimillion dollar project, and it's just unveiled. In this era of experimentation, this is very different. I talked about, "You're not unveiling the whole house at one time, you're doing little Lego blocks, little building blocks." You put one building block out there. If it works, you might build another one; if it doesn't work, you rip it out and try a new one, right? So, that is a big shift in the HR organisation.
But the second shift that's happening in the HR organisation is, many of us and many of our processes get measured on 100% compliance, no defect. We're used to rolling out programmes and maybe having a 150-page FAQ where we've thought about every single possible permutation. That's not the world we're living in either, right? So, we have to know that we're going to launch something, maybe using AI, people will say, "But it doesn't cover this". "You're right, it doesn't cover it yet. Maybe next week, maybe next month, as we kind of make the next enhancement". But this idea of minimal viable product and getting comfortable with that, getting feedback on it and evolving, I think is the next thing.
[0:44:24] David Green: Very good. What are key learnings you'd like to provide to listeners on where to get started with using AI in HR?
[0:44:31] Nickle LaMoreaux: Yeah, I think there's a couple things. And again, IBM's been on a long journey with it, but it hasn't been perfect. We've made mistakes along the way. So, if I look back over our journey on this, and if there are things I would have done differently or things that I've learned, there's a couple of them. The first one I would say is, start small. You don't have to be big bang. With some of the AI experiments we did, we did go a little too fast. We did try to be too big versus start small. And so again, the way the AI technology is working is that you can start small. You can launch a chatbot that only does two things. It doesn't have to answer your whole HR talent questions, right? You can get people used to using it, you can get some value out of it. So, don't be afraid to start small.
The second thing is, and as we've been talking about, assess the impact. There are lots of things that you can do with this AI, and some of it might have no impact, some of it might have incremental impact. So, think about, if you're going to run experiments, there's limited time and resources, what are the experiments that are going to have the biggest impact? Invite employees and managers into the experiments with you, design with them in mind. Or if you're using a digital agent that's going to transform something for an HR professional, use those HR professionals to help design the digital agent. This is not tops-down, this can be very much bottoms-up.
Counter fears. Many of us are hearing the dialogue out there about, "Is AI going to replace jobs?" Be clear. As administrative tasks free up, what do you want people to do with that time instead? How do you train to get them ready for it? And then, maybe the last thing I would say is create some advocates in the organisation. Again, think of yourself like an offering manager for a product. If you're going to try and experiment in HR, find some business leaders that might want to have this piloted in their organisation. Learn from it, have them sell it to some of the other business leaders, and that will help you kind of get the ball rolling on this.
[0:46:52] David Green: So Nickle, this is the question that we're asking all the guests on this series of the Digital HR Leaders podcast, how can HR leaders harness the power of employee insights and analytics to revolutionise the workplace experience?
[0:47:08] Nickle LaMoreaux: Well, this is also a very big question, David.
[0:47:10] David Green: It is a very big question, isn't it?
[0:47:10] Nickle LaMoreaux: We could spend several hours on this. Look, maybe I'll offer a little perspective. It might be different than some other guests. I know many people could answer this question in different ways. I think one of the big shifts, as we think about employee insights, analytics, I mean as you know, in many, many organisations, the HR professionals have been on this journey for a long time and we're going to need to continue to stay on this journey, this is not going away. But one of the shifts that I think is happening is, we've often thought about employee insights and analytics as us uncovering these insights and telling the organisation about themselves, we have too much attrition here; we have not enough capacity here. We're surfacing this and we're telling the organisation.
For me, I think that now things are being shifted. Because of AI platforms, because of what we just talked about of getting feedback from users, we now have an ability to listen and get analytics and insights from a lot more people in real time. And so, the AI is allowing us to ingest that feedback in real time. And so, I think the shift we need to make is, rather than us using analytics to tell the organisation, let the organisation tell us. And in some of our AI experiments, our digital assistant, we get thumbs up, thumbs down. People will call an employee to say, "Wait, why'd you give this a thumbs down? Why didn't it work?" It is this user centre view to get the data coming to us this way that then is the feedback loop back about how we might change.
[0:49:04] David Green: I really like that, that's fantastic, and it is a question that we probably could have spent a whole episode on, so I do appreciate that, Nickle. So, Nickle, we've come to the end of our conversation, fantastic as I knew it would be. But thank you again for joining us on the Digital HR Leaders podcast. Can you let listeners know how they can follow you on social media, maybe find out more about the work that you're doing at IBM?
[0:49:29] Nickle LaMoreaux: Yeah, absolutely. So, you can certainly go to the ibm.com website, you'll see a lot about some of our HR use cases, but I also encourage others to follow me on LinkedIn, where I often talk about experiments that we're doing and how we're continuing to use data in AI to make our function better.
[0:49:46] David Green: Fantastic, and we'll put a link to the HiRo case study, which is on the IBM site, that we talked about earlier as well, because I think it's a really good example, I think, to inspire people. Nickle, thank you so much, and I hope at some point that we'll meet in person, maybe at a conference, probably where we'll both be speaking, perhaps.
[0:50:05] Nickle LaMoreaux: I would love it, David. Thank you for having me.