How Enterprise Talent Leaders Deploy AI Coaching at Scale: A SIOP 2026 Panel

Speakers

Beth Shipman, Director of Culture & Talent Insights, Corning. Leads global talent assessment strategy and culture and employee listening at Corning, with prior experience in public safety testing and federal government training.

‍

Liz Ritterbush, Manager of Talent and Performance Management, The Home Depot. Oversees performance management programs across The Home Depot's 400,000 to 500,000 associate workforce, with a background in IO selection and assessment.

‍

Rob Lewis, VP of Talent Management and Org Effectiveness, ADP. Leads talent planning, succession, executive development, assessments, org surveys, and change management at ADP, with prior roles at PepsiCo, Pfizer, and Prudential.

‍

Julia Walsh, Manager of Talent Assessments, General Mills. Leads global talent assessment practice at General Mills, covering selection and development for every role in the company from hourly production to C-suite.

‍

Ellie Wildman (Moderator), Director of Strategic Accounts, Valence. Works with enterprise clients to deploy AI coaching, design behavior change programs, and define what good looks like across audiences.

Key Takeaways

• Adoption builds when AI coaching is embedded in existing talent processes. Corning has logged more than 30,000 unique coaching sessions in less than a year with a 56% return rate, driven by integration into culture survey debriefs and performance management.
• AI coaching applies frameworks that general AI tools do not. Liz Ritterbush at The Home Depot described testing the same workplace conflict with multiple tools: ChatGPT and Copilot acted as cheerleaders, while Nadia asked about her desired outcome and applied coaching models.
• Use cases evolve dramatically over months of use. At ADP, leaders used Nadia roughly 4x more frequently in month six than in month one, shifting from summarization tasks toward leadership capability building.
• Aggregate insights surface organizational themes traditional surveys miss. Beth Shipman at Corning identified regional stress patterns through Nadia data that aligned with concurrent culture survey trends, enabling earlier intervention.
• AI coaching scales 360-degree feedback without proportional HR investment. Rob Lewis at ADP highlighted that Nadia eliminates the need to train large HR populations in feedback delivery while preserving quality.
• Customization matters and is iterative. General Mills, Corning, and ADP have each trained Nadia on their leadership frameworks, IDPs, values, and regional resources, with Liz Ritterbush noting the work continues "a full lap around the sun."

‍

Valence's Vision: AI Coaching That Deepens Human Connection

How AI coaching differs from general-purpose AI tools

AI coaching is purpose-built to drive behavior change, applying coaching frameworks that prompt reflection rather than supply answers. Valence's Nadia builds a profile on each user, understands their team and organization, and surfaces proactive insights about colleagues and upcoming moments. Unlike general AI assistants that summarize or generate, AI coaching asks probing questions, references the user's own goals, and reinforces leadership capabilities the company has trained the coach to recognize.

‍

Ellie: Welcome, everyone. Thank you all so much for joining. I will give a brief intro, and then we'll get to the core part of why we're here, which is speaking with our fantastic panelists. Before we get started, [I'm] Ellie Wildman, [a] director here at Valence, and in my role, day to day, I work with clients to think about deploying AI coaching. So how do you actually create [00:00:30.039] that behavior change? What audiences might resonate? And then what good actually looks like.

‍

I'll give a brief intro of who Valence is, and then we'll spend the rest of the time today speaking with our panelists. Nadia is [an] AI coach. How we really think about that is at a holistic level. We [00:01:00.140] started with AI coaching, and our founders were a part of McKinsey, and they were really focused on coaching and then executive talent more broadly.

‍

What they realized was it's super beneficial and really hard to scale. What we ended up doing was essentially trying to recreate AI coaching. What we found very quickly is that that's not necessarily always what resonates [00:01:30.079] with folks. People want a mix of development, a mix of reflection, and a mix of in-the-moment and task support. So that's really how we've built Nadia.

‍

I'll talk a little bit more about that, but as we are thinking... Before I dive into Nadia more broadly, I want to talk a little bit about what we're seeing in the market today and our vision for the future. Our belief is that AI [00:02:00.180] is actually going to deepen human connection instead of replace it. First is closer teams with fewer blind spots. One of the things that Nadia does is she understands you, she understands your team, and she actually proactively raises insights that you might not recognize before.

‍

For example, my Nadia might know Liz, and I might have a meeting coming up with Liz, and Nadia will say, "Hey [00:02:30.080] Ellie, I know you're a bit conflict-averse, and this is actually how Liz receives feedback best. I actually wouldn't recommend giving Liz feedback in this way. I would actually recommend giving feedback to Liz in this way." We'll surface things that you might not already know.

‍

The second thing, [you] probably already have seen this and heard this, but this concept of HR is the new R&D. The future is filled with both agents, [00:03:00.360] people, and what that combination looks like is really on HR to help drive and to help paint that vision and make sure that the future is human-focused and really human-led. That's on us to help think about what that future looks like.

‍

The third is this idea of really contextualized support in the flow of work. Every person and every day, and I would even expand that to every moment. [00:03:30.439] It's [not] development every quarter. It's [not] development maybe every two weeks when you happen to meet with your manager. It's development in the flow of work that's really contextualized to you, your role in your organization.

‍

I won't go through all of this, but all this to say is Valence has taken a very AI-first approach on this, so you'll see this in who we have on our [00:04:00.120] team. We started with fantastic people that were focused on coaching and team effectiveness, and have brought in the world-leading experts in what we call conversational AI and AI more broadly. AI that doesn't just answer questions but AI that actually helps you reflect, and the actual systems in place to help you do that.

‍

I won't go through all the client examples because we have our fantastic panelists here to talk about that. A little bit of what Nadia is. [00:04:30.297] Our mission is to provide contextualized coaching in the flow of work. A few ways we do this: one is Nadia knows you, so she builds out a profile on you, she deeply understands you, your goals, what you're working towards. The second is she understands your team, so exactly the example earlier with Liz, and can actually surface those insights.

‍

Most relevant, I think, to this audience is Nadia knows your organization [00:05:00.620] and your talent moments. The way that I like to think about your organization is we train Nadia on three parts, and we'll hear a little bit more about this from our panelists. One is at the org level: your mission, your values, your ways of working, so that you can really scale those throughout the organization.

‍

The second is at the functional level. We like to say we understand that a manager in marketing, what success looks like, is super different than a manager in sales [00:05:30.579] or supply chain. Nadia actually has that context and can coach accordingly. The last thing is your talent moments, which I'll talk a little bit about.

‍

The concept is in 2023, we were focused on AI coaching for the individual, and coaching and support for the individual. In 2024, this shifted to how do we actually do that at the enterprise scale. It's about you [00:06:00.079] in relation to your organization. How do you support talent priorities, and how do you support the individual? Now, the future is really about this idea of teams. How do you actually support the groups of individuals that do the work?

‍

I won't spend too much time here. As we're thinking about, and as you all are reflecting [on], what does this actually look like? We often think of deploying Nadia in two ways. One is this idea [00:06:30.040] of on-demand coaching or in the flow of work. I have a call in 15 minutes, I'm really nervous, I don't know how to prepare. A great use case for Nadia. The second is this idea of talent platform, or a talent lifecycle more broadly. That would be talent moments, programs, and org change. We'll hear a little bit more about that, but hopefully, this is helpful to ground you in what Nadia is.

‍

Meet the Panel: Enterprise Talent Leaders on AI Coaching

Who is on the panel

The SIOP 2026 panel brought together four IO psychologists and talent leaders responsible for AI coaching deployments at major enterprises. Beth Shipman leads talent assessment, culture, and listening strategy at Corning. Liz Ritterbush manages performance management programs across The Home Depot's 400,000-plus associate base. Rob Lewis heads Talent Management and Org Effectiveness at ADP. Julia Walsh leads global talent assessments at General Mills. Each speaker brings a distinct enterprise context to the conversation.

‍

Ellie: With that, I'm going to start and turn it [00:07:00.100] over to our fantastic panel. Before we jump in, can everyone hear me? Before we jump in, I would love to have each of you introduce yourselves and take a couple of minutes. We have such a plethora and fantastic experience represented here. Maybe you can do a little bit of who you are, what your role is. [00:07:30.740] Who you are, what your role is, and then what brought you on that path. Then a little bit more about what your role looks like today in relation to coaching. Beth, I know that you have done a ton of work at Corning more broadly. Would love to hear a little bit more about yourself.

Beth Shipman, Corning

Beth: I started my IO journey at Radford University. Today, I'm at Corning, and over my 10 years there, I've held a bunch of different roles in talent management, center of excellence. That ranged from global leadership program management, and today, managing [00:08:30.353] our global talent assessment strategy, as well as our culture and employee listening strategy. In between Corning and Radford, I started out in the public safety testing sector, and then transitioned a little bit more into change management, training, and online tests development for the federal government.

‍

Coaching, for today: I help a lot of our HR business [00:09:00.259] partners match the right type of assessment to the right need, and that looks differently across Corning's multinationals. We have a ton of different requests and needs, and we're trying to put together a scalable strategy. Some of that looks at 360s, coaching, executive level through frontline, and Nadia is now part of that solution as well.

‍

Ellie: Awesome. Thank you, Beth. As you're going through [00:09:30.179] too, Beth, it might be helpful to hear a little bit more about what is the biggest problems that are on your mind right now? What are the biggest problems that you're thinking through in your current role?

‍

Beth: I think scalability is certainly one of the problems. Right-sizing that in the world that we live in today is a different challenge that we felt compared to 10 years ago, or even 2 years ago, in trying to really live our employee [00:10:00.440] value prop across the entire globe really consistently at scale, but in an authentic way.

‍

From an AI perspective, because I think that's why we're all here, really understanding how that's going to impact relationships at work, making us better versus the messy middle that we may be in right now, figuring that out, and organizational culture impacts.

Julia Walsh, General Mills

Ellie: So powerful. Julia, I know you and Beth go way back. I think [00:10:30.039] we were on a panel last year, and one 15 minutes ago. Would love to hear a little bit more about your role at General Mills, and how you have grown into that role, and some of the biggest problems that you're seeing.

‍

Julia: Hi, I'm Julia Walsh. I've [00:11:00.539] been with General Mills for a little over two years now, and I lead our global talent assessment practice. That looks like selection, it looks like development, it's for every single role in the company globally. Everything from our hourly production workers all the way up to our C-suite. It's exciting, it's a lot, and now that we're rolling things out a little bit more consistently globally, there's additional challenges there. [00:11:30.679]

‍

Related to AI and coaching, for being a 100-plus-year-old, fairly traditional company, we are really leaning into AI, building AI and digital fluency at every role at every level. AI coaching is a part of that. While I would say we understandably so invest a lot into our executive-level coaching, and it's much more hands-on, white-glove, this and [00:12:00.539] tools like this, Valence, are a perfect fit for everybody else, democratizing coaching and having it just always on, always available. I can speak to it later, but just really embedding it into our processes has been a huge, huge win for us.

‍

Ellie: General Mills was quite early in terms of embedding it into their systems and programs, and you all have been quite thoughtful about that, so we [00:12:30.039] look forward to hearing that.

Liz Ritterbush, The Home Depot

Ellie: Liz, I have the absolute pleasure of working with Liz closely. You represent a huge workforce. Would love to hear a little bit more about your role, how you got there, and some of the challenges that you're facing.

‍

Liz: Hi, everyone. I think I see some familiar faces in the crowd. I'm Liz Ritterbush. I am a manager of talent and performance management at The Home Depot. Any time [00:13:00.100] during the year, we're anywhere from 400,000 to 500,000 associates and employees across the organization, and across the country. We're also in a few other countries as well.

‍

I started traditional IO in the selection space. I started with assessments and then moved into performance management, focusing on goals. What we're looking at is how can we leverage AI to help with the performance management piece, to help with the conversations that leaders are having with their [00:13:30.065] associates? Help to filter out all of the noise, because in an organization that's half a million strong, you've got a lot of voices coming at you, so how do you filter those through? We're looking at different ways to incorporate AI, we're testing on a lot of different things, and AI coaching is one of those, for sure.

Rob Lewis, ADP

Ellie: Last but certainly not least, Rob. You've been deep in IO. When we were first talking, we were all sharing [00:14:00.080] how many years we've been coming to SIOP. I won't put you on the spot to tell everyone how many years, but it was quite a bit. You've also been deep in IO roles at several different organizations. We'd love to hear a little bit more about your role right now, and what led you on that path.

‍

Rob: My name's Rob Lewis. I have been in the talent and development space at various companies for about 20 years, big companies like PepsiCo, Pfizer, Prudential, now ADP.

‍

Today at ADP, I head up a group called Talent Management and Org Effectiveness. My group is responsible for the talent planning process, succession, executive development and coaching, assessments and feedback process, as well as our big org surveys, [00:15:00.440] org design, and change management, a lot of IO stuff.

‍

I've been with ADP about six years now, joined during the early stages of COVID, March 2020, but it's been a good journey there. In terms of what's unique at ADP, we have this mix from a people leadership standpoint of really long tenure leaders who've been there 10, 20 years even, and those folks are looking for new development that they haven't [00:15:30.179] had in a while. "I haven't had exposure to a program in a long time, I don't get new learnings in my role."

‍

Then we have a lot of new people leaders. About 1,000 leaders are promoted into leadership roles every year at the company. A lot of people are just coming up to speed learning how to be a leader. Those people also need a lot of learning, and we do executive human coaching at senior levels, but as we've said earlier, that's not scalable, [00:16:00.340] so we've really leaned into AI coaching as a way of meeting the needs of our diverse people leadership population in a scalable way.

‍

From Skeptics to Believers: IO Psychologists on AI Coaching

Why IO psychologists became believers in AI coaching

IO psychologists who initially doubted AI coaching reported shifting their view after experiencing built-in coaching frameworks like the GROW model in practice. Beth Shipman at Corning described the discovery process where Nadia learned the composition of her team and used genuine listening behaviors. The platform's coaching scaffolding, including reflective questions and follow-up nudges, distinguished it from general AI tools and delivered the kind of pattern recognition skeptics expected from a trained human coach.

‍

Ellie: As you can tell, a very impressive panel today. We'll start. I know there's a lot of IOs in the audience. Beth, I'm curious, you self-described as a skeptic at first. I'm curious of [00:16:30.480] how your mind changed throughout the process, what changed your mind, and then why are you still skeptical now?

‍

Beth: Working with a bunch of scientists and engineers and having married an engineer, I come from that side of personality. We did a pilot to start out, it was about a year ago, and we started with our science and engineering organization. We knew, "Let's try to win over the skeptics," and I was a part [00:17:00.009] of the pilot too: "Yeah, let's break this thing, humans are going to dominate coaching, how can you replace that?" Different solution, but as we went into that journey, the pilot, we got some really amazing feedback. Me personally experiencing that, I was able to really represent the IO side up front in the design and know where we may want to further influence and nudge and embed parts of what makes Corning Corning, our values, things like that.

‍

In the initial discovery portion, you're asked some intake questions, [00:17:30.339] just like in human coaching. We'd probably fill out an intake form: "Tell us a little bit about your role." Very quickly, Nadia started to learn the composition of my team, and whatever I was willing to put in. That context really showing that it was listening, and you think about good coaching behavior, and [00:18:00.119] you really want someone who listens and picks up on that nuance.

‍

That process really started to win me over, when I realized there's some real coaching framework built into this experience, and I could see that, but the average person may not know that the GROW model is here. Just those little proof points: turning a skeptic into a believer is in what you see. That process for me was organic, and I'm [00:18:30.259] glad I was a part of it up front, to be able to address some of those things, when I was asked what's this whole thing about.

‍

Ellie: That's so powerful, especially your own journey, coming from a skeptic to less of a skeptic. I'm curious, you all have rolled out Nadia to a huge population. How do you think about getting the rest of the skeptics on board? How has that process been?

‍

Beth: Right now, [00:19:00.160] I think we're at like a 56% return rate. When people see it, they come back, and that's a good proof point, and I think they're telling each other too. It's not me having to sell it, it's figuring out when do we point people to it. I think embedding it in processes helps, like if someone doesn't know where to go, single sign-on's great, but if I don't know where to go, I may not discover it. Embedding it into our processes has just naturally started to grow the user base. [If they] tried it out, [00:19:30.493] they're more likely to come back, I think, is our learning.

‍

So far, I think we're doing well for adoption. How else can you get... I think we're over 30,000 unique coaching sessions right now in less than a year. You can't do that at scale, having good frameworks behind it, and really having someone... You dig below the surface, I guess, is how I describe the experience. It's not telling me an answer. [00:20:00.440] Good coaching is asking a question back, and having that person uncover it and learn.

How AI coaching distinguishes itself from general AI assistants

Liz Ritterbush, Manager of Talent and Performance Management at The Home Depot, described a personal test where she brought the same workplace conflict to ChatGPT, Gemini, and Nadia. ChatGPT and Copilot acted as cheerleaders, encouraging her to confront the colleague. Nadia paused and asked, "What is the outcome of the conversation that we want to have?" That moment, applying coaching models in real time, marked her turning point from skeptic to believer.

‍

Ellie: I'm curious, the rest of the panelists, if anyone else has either shifted from a skeptic to less of a skeptic, or had a specific audience that they were struggling with onboarding and getting that behavior change in your organization, that they have now are promoters.

‍

Liz: I'll lean in here, [00:20:30.019] as someone who also married an engineer. For this, I was a skeptic going in, and I still will challenge it, I'll still test it. I still encourage other people to challenge and test the tools. Anything that we're looking to roll out, let's see how we can refine it. What I love, and I was going to get into this a little bit later, but I might get into this now: I had one of those days, and I'm sure we've all had those days, [00:21:00.019] where you have a conversation, you have something at work that's just on your mind, and it's on your mind in the evening, and it's on your mind the next day when you wake up.

‍

I woke up, and it's still on my mind, and I pull out the laptop, and I start checking with the AI coach, and I'm checking with ChatGPT, and I'm checking with Gemini too, because I'm a skeptic. I'm telling it, "I'm just frustrated, I don't understand, it's been a week of not being able to move [00:21:30.240] this one person, and I'm ready to confront them. I'm ready to go out and tell them how I feel."

‍

The AI coach was like, "Hmm." Going back and saying, "Well, is that the smartest idea? That might not get you the outcome you want." ChatGPT and Copilot were my cheerleaders. They were like, "Yeah, go out. Go challenge that person. Let us know what resources can I give you?" Nadia was like, "Let me see. What is the outcome of the conversation that we want to have?" [00:22:00.740] What do you want to accomplish in that conversation? That really was a turning point for me with challenging my skepticism, because I was like, "Okay, it's actually using coaching models just like a professional leadership coach would."

‍

Ellie: Sorry, go ahead, Julia.

‍

Julia: I was just going to add that, personally, I'm more of an early adopter, so I personally wasn't a skeptic, yet [00:22:30.099] I did see people around me, so it was more like that empathy building to understand what are the barriers to entry for others. Because I'm all on board. I was in it, I was messing with it. I was asking it things, my friends were too. It was more about going outside of my little circle of early adopters, and really hearing from other people, what are the barriers to entry?

‍

Part of it, it was just lack of awareness. They just didn't know its power, [00:23:00.000] and what it could do, and what it couldn't do. That was part of it. I'm also going to allude to things that I might say later too, but really embedding it into processes that we already have. We have an individual development planning season every spring, and it was the perfect opportunity to say, "Are you stuck building your IDP plan?" I know I was, and honestly, my manager had to remind me, "Go to Nadia. She's been trained on [00:23:30.220] this."

‍

We were able to train her internally on our framework. I just put in, I was dreading it, "Oh, I don't want to talk about myself. I don't want to talk about my development plan. Ugh, this sounds gross." But I went in, and I was able to just say, "Help me build my IDP plan. I want to talk to my manager about it." Something really simple, and she started very easy, asking me questions, probing questions, trying to get to the heart of [it]. It wasn't [00:24:00.019] "Let's just start building it," or "I'm going to build it for you." It was very much, "What are your goals? What are your ambitions? What things do you feel like you could develop?" She was asking really thoughtful, deep questions, and it was so much easier to go down that path than just stare at that blank sheet of paper.

‍

Finding those opportunities to remind folks, "Hey, we have this tool available, it's available to all of you, and here's how to start," was the biggest thing that helped us just get people [00:24:30.259] in there. Once they get in there, then just like you, they continue using it.

‍

Building User Judgment: AI Coaching vs General AI Tools

How to help users choose the right AI tool for the right task

Building user judgment requires explicit guidance on which AI tool to use for which purpose. Corning, General Mills, and other enterprises run AI literacy programs and publish governance over personal data, distinguishing AI coaching for reflection and capability building from general AI tools for summarization or content generation. Beth Shipman noted that coaching outcomes, role plays, and growth conversations require purpose-built AI coaching, while Julia Walsh emphasized clear guidelines on what personal data is appropriate to share with each tool.

‍

Ellie: I think that's so powerful, this idea of how do you flag... How do you help create that behavior change in users? Liz, I'm curious, you touched a little bit on this, but how do you think about building judgment of users, of when to use AI, when to go to coaches like Nadia?

‍

Liz: That's something that we're struggling with right now, because [00:25:00.160] I'm sure we've all seen it: people go into AI, or they go into their Google platform and they do a search, and they get a paragraph back and they're like, "Okay, let me trust this." But I alluded to this a moment ago: those models are going to tell you you're always right. They're going to validate whatever you say, and you're going to see more polarization with those models, unless you've trained them explicitly not to, which, good luck keeping them doing that.

‍

But like a coach, a real good AI coach, or a person coach, they're going to challenge [00:25:30.240] you on those things. They're going to have you ask questions. We're not seeing our users on ChatGPT, or on Gemini, or on Copilot. We're not seeing them actually challenge the AI. We are seeing some of that challenging, some of that dynamic, some of that conversation happening with the AI coaching models, but the coach has it built in, which is one thing that I love. It's encouraging this judgment from the user that we need, frankly. [00:26:00.360]

‍

Ellie: I think this idea of judgment, which we've talked a little bit about, especially earlier, is this idea of how does AI actually enable and teach judgment, versus taking that away from users. That's a great point. Before I transition, Beth, Julia, Rob, anything to add to this idea of judgment, and how we ensure that users are going to Nadia, and tools like Nadia in the right ways? [00:26:30.317]

‍

Beth: Our HR function is partnering with our IT organization to really accelerate adoption. Every company is trying to do this right now. The question on everyone's minds: which tool do I use for which purpose? We're trying to accelerate that through some targeted learning series. It's about a four-week series, and there's a theme every [00:27:00.192] week, and we're encouraging our entire function to be early adopters, and then to model that throughout the organization.

‍

You have to start somewhere, and it's live examples, it's demonstrations, and it's building literacy, and fit for purpose, not just more, more, more, because that doesn't actually solve the problem. Next week, [I'm going] to be leading one of those sessions for the AI literacy program, and it'll be my [00:27:30.640] own experiences using Nadia to accelerate my IO consulting process internally. That should be a fun session.

‍

Just showing success stories and good uses. When you need a human interaction, a role play, or a coach for growth and development, you're not going to get the same experience when you go to a different AI tool compared to what you're going to get with Nadia. That's [00:28:00.359] really what I've seen, and hopefully, the example I'll share next week will give some tangible scenario to others to try out too.

‍

Ellie: So powerful. Sorry, Julia, were you going to say anything before I transition?

‍

Julia: Just one more thing to add that you triggered a thought in my mind. We have advice on what tool to use when, because we have multiple AI tools, and it's a big question. We also have governance over lots [00:28:30.140] of AI governance. One of the things that we're careful to message out is a lot of people, [I'm] thinking about the people who are skeptical or reluctant to use it: they're worried about their personal data. They're worried about putting very personal things into an AI tool. Who's going to see it? How are they going to use it?

‍

We're very clear about the guidelines on what types of personal data they should and should not put in there. We do iLead assessments. While we encourage [00:29:00.119] them to use it alongside their 360 assessment results, for example, or their development reports, if they're a new hire, we also want to be very careful for them not to just put their report in there and say, "Go," because it is very personal. But if they want to ask, "I got feedback on my communication skills. Can we talk about how I can further develop that?" Perfect use case. You can use your reports and the data that you get and the feedback that you get to then accelerate [00:29:30.359] your coaching.

‍

Ellie: So powerful. I'm getting the signal I should keep my microphone closer, so please let me know in the back if you can't hear me. To Julia's point, one thing that General Mills has done quite well is this idea of identifying where Nadia sits in the broader ecosystem of AI, and that will be something that all companies and organizations will have to consider moving forward.

‍

Data and Adoption Insights at Enterprise Scale

What aggregate AI coaching data reveals about a workforce

Aggregate AI coaching data surfaces themes that traditional engagement surveys can miss. Beth Shipman at Corning observed that during high-growth periods in certain regions, themes of prioritization, organization of work, and delegation began trending in Nadia conversations and aligned with rising stress signals in the culture survey. Rob Lewis at ADP uses aggregate Nadia data to identify which leadership capabilities, including change leadership, collaboration, and growing direct reports, leaders are actively working on.

‍

Ellie: As we're thinking about and transitioning a bit, one of the things that we hear [00:30:00.099] a lot from leaders is that there's so much data available now, but how am I actually using this well, and how am I using this to actually have insights and make decisions? Julia, maybe if we want to start with you, given your background in assessments, I'm curious if having more data, especially insights from Nadia, has changed what you measure and how you think about making those decisions.

‍

Julia: A hundred percent. I'm fortunate enough [00:30:30.380] to be technically on the Workforce Insights team. I work on a team where it's me leading assessments work, but we also have our listening, engagement survey work, and deep workforce insights and analytics. It's all IO psychologists, which is nice too. We speak the same language. They understand, and they ask great questions.

‍

Thinking about the data that we have... All of us [00:31:00.266] are like, "More data is better," of course, and we need to be asking the right questions. We could go off just as researchers and say everything's a cool question, but we're always keeping our business objectives in mind. We want to ask thoughtful questions that will help the business and help our business outcomes.

‍

Coming together and thinking, "Okay, what available data do we have? And what are the right questions to dig into and prioritize, [00:31:30.579] versus just going after the shiny, cool question that we just find intrinsically interesting?" I think we do a good job of finding that balance. It perpetuates within the business, then they see the value of it too. If we're answering questions that they find interesting and they find helpful, then they continue to lean into our work and say, "What else could you do? What other research study? What other data do we have [00:32:00.150] that we can look into more?"

‍

Ellie: So valuable. Others, please feel free to add on. Beth, specifically, I know you are at the intersection of culture, talent, and insights. How have you used either AI coaching or insights more broadly in your role right now? How has that shifted in the past few months?

‍

Beth: I looked at this early on as, ooh, data. Very similarly, we have some very strong data privacy [00:32:30.059] standards in place. Everything that goes into Nadia is just for that person. We get aggregate information by country, so we understand usage, we understand themes. Any other fun question, "Hey, is this sentiment trending right now or not?" we can work with the Valence team to get that just-in-time information.

‍

Over the last year, I would say informing some of our L&D programs and strategy, [00:33:00.180] we've looked at what are some of the things that we're seeing, sentiment and topics, in aggregate. We can also see within region. Is that becoming a larger theme in one region versus another?

‍

We did look at culture data as well, in terms of country-level users and engagement scores. Specifically why we implemented this was really to try to improve manager effectiveness [00:33:30.599]. Not just relying on the quality of your manager, but having great development conversations and having a coach 24/7, regardless of language, putting that in place.

‍

We did see some interesting scores improve where we had higher usage in countries. We've been able to see some value already. It's bringing additional ideas, like, "Where do we go from here?"

‍

Ellie: So many questions off of that, but first of all, anything interesting? [00:34:00.680]

‍

Beth: All of the data is interesting. We were going into high-growth mode, and we started to see some of that in our culture survey impact, manager effectiveness scores. When you have high volume in certain regions, that's really putting a lot more pressure on managers, on the HR processes. The experience that you're creating when you've got some turnover and a ton of new people coming in, you have a little bit less time to do everything, because it's high [00:34:30.039] volume across the board.

‍

We were seeing some scores trend in our culture survey related to stress, so my spidey sense started tingling, and I started to look a little more closely at some of the themes in Nadia, and sure enough, we saw some changes as we were thinking about prioritization as a theme, organization of work, delegation, things like that. We started to see trending in the region [00:35:00.079] that I would hope that we would see that trend happening, where we are able to see that support, that network that's there through the coaching tool, versus only relying when someone would speak up.

‍

Ellie: That's so powerful. Last question, and then I would love to hear from the rest of the panel too, a little bit more about the data that you're using now, and the data that you're hoping to use moving forward. Beth, as you're thinking about... I know you mentioned what you're hoping to do, and what you're hoping to measure moving forward. Is there [00:35:30.000] anything in particular that you're curious about or excited about?

‍

Beth: We're excited about trying to embed this into more of our leadership programs. We're launching a few new solutions and trying to figure that out, so I'm excited about that. Will we start to see bigger improvements on some of our culture survey scores as a next step? That, and just continued adoption. It sounds so basic, but [00:36:00.320] you love to see that scale and that consistency of people having something to work with and not just relying on me, or HRBPs or their manager. I think that's probably where we're going, understanding that intersection of culture and engagement.

‍

Ellie: Absolutely. Any other insights on data before we transition?

How AI coaching reveals leadership capability gaps

AI coaching usage data shows which leadership capabilities people actually choose to work on, beyond what they self-report. Rob Lewis at ADP rolled out a new leadership model and used Nadia data to see the top capabilities leaders were addressing in conversations: change leadership, collaboration, and growing the people on their team. This passive, behavior-based signal tells a different story than survey self-reports about development needs.

‍

Rob: It's a little bit more basic, but [00:36:30.380] we just rolled out a new leadership model about a year ago. We were really interested... We were really wanting to reinforce the importance of these certain leadership behaviors, but we also wanted to understand really intelligently what are people needing to work on? Where are the big opportunity areas that our leaders are reporting? We've asked them directly, "Okay, what do you think you need to work on?" People said certain things.

‍

It's a whole different story [00:37:00.239] when you can actually look at the data of how people are using an AI coach. We can look at the aggregate of the capabilities they're focused on. Are they working on change leadership, collaboration? Are they working on growing their individual people on their team? Those happen to be the top three.

‍

We're just starting to scratch the surface of what we can look at from a data perspective, but it's pretty illuminating. Compared to other data sources where it's [00:37:30.039] usually self-report, you're not reporting on the actions that people are taking for their own personal development when nobody is actually really listening in on them other than being able to aggregate the data. That was a pretty interesting start of the journey, I would say.

‍

Ellie: I love that too, Rob. The point around, this is not just a snapshot in time. This is ongoing data that the more users are using it, and they're using it for their real challenges, [00:38:00.360] you're not just getting 200 characters. You're getting real insights.

How AI coaching usage evolves over time

AI coaching usage patterns evolve significantly over the first six months of deployment. At ADP, an initial pilot of approximately 250 leaders shifted from basic tasks like summarization and email drafting toward leadership capability building. Rob Lewis reported that users in month six were engaging with Nadia roughly 4x more frequently than in month one or two, and the depth of use cases (preparing for difficult conversations, reinforcing program learnings) expanded as peer stories spread.

‍

Ellie: Rob, would love to dive into a bit more of ADP's journey. When we initially talked, a lot of users were using it more like a typical chatbot. They were going for summarizing. They were going for "Help me draft this email." Then there was a huge shift in leadership. [00:38:30.440] Walk us through a little bit that shift and what you think caused that.

‍

Rob: We had an initial pilot group of about 250 leaders, people leaders of different levels. We were really encouraging them to use Nadia, the AI coach, in different ways. The initial use case is that people were reporting out to us in our group chat. There was a kind of Webex-Teams as our tool at ADP. We had a Webex-Teams group chat, and we had [00:39:00.059] a couple of facilitators co-leading the journey. People were saying, "I use Nadia in this way. I summarized a bunch of text," or "I wrote an email to my team using Nadia." We were like, "Well, that's great."

‍

What Nadia is particularly good at is helping you prepare for difficult conversations or doing this kind of work. We had to start providing those nudges, and over time, people started reporting back, "You know what, I tried it out, and this was even more valuable. [00:39:30.260] This was something I couldn't do with other tools. This is something that really helped me actually build my capability as a leader."

‍

People started coming back with their own stories, and those stories started snowballing a little bit. That's when people started getting the most value out of it because it wasn't initially clear, but once people heard the stories from their peers, they heard other ways, and they started experimenting. They got a little bit more comfortable. [00:40:00.480] That's when they started opening and unlocking the power of the tool over time.

‍

Ellie: So powerful. Continuously coming back. Your first conversation is so different than your 10th or your 100th. Beth, Julia, Liz, I'm curious if you had similar arcs at your organization.

‍

Julia: Yeah, for sure. Going [00:40:30.019] back to what Beth said, we definitely make it very clear we're using aggregate data. In the conversations we're having, once we get over that initial, "Should I be doing this? I'm nervous." Once we get people in there and trying it, having those conversations and sharing different use cases in breakout rooms and saying, "How did you get value out of it?" "Well, I tried this thing." "Oh, I didn't know you could build a persona in that and have [00:41:00.079] a role play." Having those conversations.

‍

They happened fairly organically when we just start prompting people, and then they get ideas, and they say, "Well, I wonder if I could use it for this conversation that I've been dreading having with a coworker." Then it really does just build on itself. We do encourage those conversations. We set them up so that they're happening. We'll even do it. We'll all be in a Teams meeting, and we'll put in [00:41:30.019] the chat, "How have you used AI coaching or Valence?" Or "How have you used Mill's chat?" which is our AI chatbot. We're always prompting, and we're doing it in informal ways and formal ways, so that people are sharing ideas, and it really does build off one another.

‍

Ellie: So powerful. As we think about behavior change [00:42:00.079] more broadly, what does that actually look like? How do we make sure users are using it in the flow of work? As we're thinking about behavior change more broadly, I'm curious, Rob, as you're thinking about ADP and your journey specifically, were there any things that surprised you in that arc?

‍

Rob: What evolved was not just that people were going [00:42:30.199] into the tool more because actually, they were starting to use it more. I think it was probably a 4x change. Over the first six months, we had users using the tool four times more frequently in month six than they did in month one or two. It was not just how frequently they were using it, but they were evolving how they were using the tool.

‍

When we were looking at what the tool told us people were doing as well as what people were reporting out, it was much more on the "how can I build my own leadership capability" [00:43:00.320] kinds of functionality. It wasn't just the more general use cases. That ended up being really powerful to the point where we're doing a lot of the same things that were shared earlier in terms of building this tool into our leadership development programs. That's where people have told us, "This is a tool I want to see more of."

‍

We had a new leadership development program where we had various ways of learning. We had an in-person [00:43:30.539] learning discussion piece of it. There was a self-guided e-learning. There were short videos. We had Nadia, the AI coach, to reinforce the learning. We asked about all the different elements. Frankly, the AI coach got the highest marks because that allowed the learning to be super personalized.

‍

It wasn't just a one-size-fits-all, "Hey, this is what everybody needs to know about change management or this leadership capability." It was, "What do you [00:44:00.000] specifically need to work on?" It already knew a little bit about you and what you needed to work on. Now it knew that you were exposed to this new content that it had to reinforce. The combination of those things came together in a really powerful way.

‍

Embedding AI Coaching in Talent Programs and Moments

How AI coaching integrates into performance management

AI coaching integrates into performance management by supporting both the conversations leaders need to have and the policy navigation employees raise. Liz Ritterbush at The Home Depot rolled out Nadia in tandem with a new performance management system, and observed users initially treating it as a search tool, then evolving toward asking how to navigate policy and prepare for difficult conversations. Users effectively treated Nadia as an HR partner available outside business hours, which relieved pressure on HR business partners.

‍

Ellie: Super powerful. Thank you, Rob. I also think what you brought up of integrating it into moments and programs is so powerful. Before we shift, I'm curious on how others, Beth, Julia, and Liz, have thought about integrating Nadia into these moments [00:44:30.139] and programs and how that's allowed you to reimagine what's possible for learning.

‍

Liz: We actually started pilot testing AI coaching because we were changing our performance management system. We rolled them out in tandem. The transitions we saw were people were using them as a search tool for the first month, and then they started realizing, "Wait, I can ask it more questions about the performance management system and I can ask it about these conversations, [00:45:00.400] and I can ask it about this."

‍

We saw that evolution that you all are talking about naturally. Those case studies were also really helpful for helping us talk to other stakeholders within the business that weren't necessarily bought in. We were able to show them the journeys, not specifics, but general journeys of how people were using the AI coach. I think... I'm having a brain lapse here. Go back to... What was the question?

‍

Ellie: The pace and performance [00:45:30.380] management.

‍

Liz: Performance management. Really, embedding it, rolling it out as part of, "Hey, there's this new system," and "Hey, there's this tool to help you," was very helpful for our users, and it made them feel supported in a time, in an age where everything's changing constantly and they're constantly overwhelmed.

‍

This helped to reduce that cognitive overload that they're all feeling and that everyone in this room is probably feeling, [00:46:00.300] right? It helped to synthesize some of that information for them and helped to make that transition a lot easier. If I were to do it again, I would again incorporate it into something that they were nervous about when I roll out.

‍

Ellie: That's so powerful. I'm curious, before we transition, Liz, is there anything that surprised you of integrating it into performance management?

‍

Liz: What surprised me is how much they used it as their HR partner. [00:46:30.139] It became their HR partner. Their HR partners were relieved, but [users] asking a lot of HR what is the policy around this? How do I navigate this policy and still do what's right for my employees? How do I have this conversation, knowing that there's policies in place? We thought it was going to be more intrinsic coaching, but it was really, "How do I manage these systems? [00:47:00.239] How do I navigate these systems?"

‍

Ellie: That's so powerful too. I know we've talked a little bit about this, but there's this idea of, we're so focused on the programs and learning programs that we're doing, which are critical, and a user has 1,000 other things on the top of their mind every single day. How do you have a tool for all of those things? Before I transition, Julia, Beth, any other programs or moments [00:47:30.460] that you all have considered or the reactions there?

How AI coaching supports individual development planning

AI coaching supports individual development planning by walking users through goal-setting questions that overcome the blank-page problem. Julia Walsh at General Mills described using Nadia, trained on the company's IDP framework, to build her own development plan during the spring planning season. Nadia opened with questions about goals, ambitions, and growth areas before any plan content, making the conversation easier than starting from a blank template.

‍

Julia: I think our team who helps lead some of this work does a great job of always looking at and asking the question: are there opportunities to mention Nadia? Does it make sense to mention Nadia when we send out the support materials for our 360 assessment? Does it make sense to mention Nadia for our onboarding development reports? They're always asking.

‍

Sometimes the answer is, [00:48:00.039] "Maybe not the right tool," but a lot of the times it's, "Oh yeah, I hadn't thought about that. Let me just add a blurb at the end of the email saying, 'And if you want coaching on this, go to go/nadia,'" and it takes them right there. Just asking the question a lot to say, "Does it make sense to put it in here?" A lot of the times, the answer is yes, and then it just makes it so much easier for somebody to think of it. Often it's just they might not think about it. They might not be aware that that tool can help them in that moment. [00:48:30.380] So there's that piece.

‍

The other thing, this probably goes beyond the scope of this conversation, but thinking about team development resources and perspectives, and align. Those have been hugely successful. These are other tools in the Valence ecosystem, but we have used those in some cases, and that reminds them, oh, and there's a coach that you can use as a follow-up. Sometimes just opening that gateway has been a good opportunity as a follow-up [00:49:00.039] step.

‍

Ellie: Now, actually, Julia, we would love to hear a little bit more about that because now Nadia knows your perspective, Nadia knows a bit more about you, and can help with that team development aspect as well. Would love to hear the reactions from General Mills thus far.

‍

Julia: It's been really positive. From what I've heard, and I've experienced it myself, I've also heard from other leaders who've used it, the expectation... The only bump in that is we don't have [00:49:30.019] a dedicated team to lead those sessions, yet once you do it, it's very self-serve. It walks you through all the steps, so you don't have to be a trained HR professional or a trained assessment professional to go through those steps. The way that those tools are built, there's anonymous feedback and comments that are options.

‍

It really does build that psychological safety in the team to have an outlet to give feedback to a manager or to teammates [00:50:00.139] to say, "Hey, here's what I feel like is working well in the team, here's what I feel like is not working well." It really is a good thought starter, discussion starter. We've had great success, from what I understand, in teams that have used that.

‍

Ellie: I do think it's this idea, Julia, exactly to your point of we all, as IOs, know the value of coaching frameworks, of all of these things. Your everyday manager might not, likely will not, so how do you get support? How do you get coaching in the hands of [00:51:00.000] managers, ICs, everyone at your organization, in a way that will resonate with them and is also helpful for them? That's super powerful. Beth, I want to make sure you have space to talk about any programs or moments that you all have integrated Nadia into.

How AI coaching helps managers act on culture survey results

AI coaching helps managers act on culture survey results by making best-practice debrief guidance accessible at the moment results land. Beth Shipman at Corning described embedding a customized version of Nadia in every manager's results email, replacing the traditional one-pagers, SharePoint sites, and webinars that managers rarely consulted. The framework was already there. The change was meeting managers where they were, with reflection prompts and nudges to take a single, simplified action.

‍

Beth: For us, it was perfect timing to incorporate it into performance management. It just went hand-in-hand. That was stemming from some of the feedback we got from our culture survey, [00:51:30.119] so it was a good match, and the season aligned well, so I think that helped us a lot. Not everyone is great at giving feedback, so really building that skill set pays off in many ways. Even for the person to prepare for that conversation as well. Not everyone has their HR business partner sitting in the same location, or has an IO psychologist to reach out to [00:52:00.079] to understand how should I think about my career path.

‍

Audience Member: We all should.

‍

Beth: Well, that's a little creepy moment for me. At one point, we had embedded Nadia into our culture survey process, and we soft-launched because there was some skepticism, as you can imagine. Should this really be led by HR going forward? Is this something we should expect every manager of every team? Who is in charge of taking action after these [00:52:30.159] surveys? A question we can never just put to bed. It's everyone, it's yes.

‍

Getting the best practices in the hands of our managers is what we tested. It's not that we didn't have the framework, it's not that we didn't build the one-pagers and translate them in all the languages and make them available, or have a webinar with tips. All of these things were high lift every year. Nothing changed. Putting this link to a customized [00:53:00.079] version of Nadia in every manager's email telling them their results are ready really showed us how managers can be taken to that best practice without having to go, "Where's the SharePoint site? Where do I find the document? Is it ready and available? Do I know the right process?"

‍

Just having that guidance fed naturally into the discussion, and I was able to say, "Hey, listen, we really want managers of teams to take one [00:53:30.239] action. So try to nudge, let's get Nadia to nudge for simplification, and show progress." That was probably one of the success stories, I would say, and we won some people over with that as well. Managers weren't interacting with Nadia in a defensive way, which was a little bit of the fear. We were like, "They have the data, are they going to do something with it? Are they going to have the right mindset?" That sentiment from Nadia helped us understand managers [00:54:00.340] are really trying to improve at all levels.

‍

Ellie: I love that example. Employee engagement surveys are a great example too, because you get these results back. It's emotional. You don't know how to debrief or reflect on the results. How to actually have that conversation with your team or whoever you need to have it with is a completely separate thing.

‍

Beth: I didn't mention my creepy moment, sorry, I totally forgot.

‍

Ellie: Yes, please.

‍

Beth: I'm sitting in a room with HR execs, and they said, "Oh, [00:54:30.059] this is really great, the customization for the culture survey. It's like I have Beth here 24/7." I was like, "I don't know if that's a good or bad thing," but it's an example of, IO is not going away, but we're able to feed the framework and the thought process to scale that. I was like, "You know what, it's serving the outcome." It's [not] a Beth, it's a Nadia.

‍

Ellie: Exactly. It's like having a Beth next to you that has all...

‍

Customizing AI Coaching to Organizational Context

How organizations train AI coaching on their culture and frameworks

Organizations train AI coaching on their context through a layered process: corporate values, leadership frameworks, performance management systems, and locale-specific resources. At Corning, Beth Shipman worked with the Valence product team to integrate the company's "three plus one" culture survey framework so Nadia would emphasize reflection first and respect a manager's locus of control. ADP integrated its leadership capabilities and critical conversation model. General Mills trained Nadia on its engaging leader framework and IDP worksheets.

‍

Rob: We've got [crosstalk] question here.

‍

Ellie: I'm so sorry, I didn't even see you.

‍

Paul (audience member): How much context, because in AI, I know that software teams spend so much time on context, and giving Claude all kinds of stuff about what we're trying to do. What was that like? Because it works really well if you do that, but I was just curious about when you onboarded it and got started to use it, just curious about what that looked like, and how much work was that?

‍

Audience: Could you guys state the question? Because otherwise we're [inaudible].

‍

Ellie: [00:55:30.559] Fantastic question. It's one about context, this idea of how do we, not only train Nadia, but what context is Nadia pulling from, which I'm happy to take a first pass.

‍

Paul: I was thinking about...

‍

Ellie: Then onboarding.

‍

Rob: From their own company.

‍

Paul: How do they let it know about their organization, their policies, their values, all that sort of stuff.

‍

Ellie: I will actually turn it over to the panel, and would love to hear a little bit more about, one, what [00:56:00.099] was that process like, getting Nadia integrated? Two, how have you seen a change in the behavior of Nadia based on the context that she has?

‍

Beth: I just gave that example, so for me, the first part of the integration, it's really like, what kind of company is this? What are your corporate values if you have them? That's the easier stuff, that level of context. [00:56:30.099] For me, I was fearful, surprised, "Is this thing going to try to nudge a manager to try to act on something that's super systemic, versus something that's within their team, and their locus of control?" That is where we had the really great discussions with the product team, to say, "Hey, we have this framework, it's called three plus one. Three global, one local."

‍

We were able to describe it in that way, to say, "We want to emphasize, Nadia's not going to stop you from [00:57:00.239] having an action plan that may or may not be fully in your control, but she'll help nudge and reinforce, and start with reflection." We wanted reflection first, not just, "Hey, a GPT can tell me, 'Here's five things you can do.'" We didn't want that experience.

‍

Those were some of the discussions we had. What's our process? What's our framework? We were able to even share some of the action plan ideas and the structure of our culture survey, defining those categories, and trying [00:57:30.340] to reinforce even the internal resources that we wanted managers to know about. It really was a robust process for us, for that example, and it took a few weeks. This was not something that... We went through rounds and rounds of testing, but I didn't need to tell Nadia what engagement was. I was like, "How are you defining employee engagement?" Some of the stuff was already there and available.

‍

Rob: We've done a pretty light integration, I [00:58:00.019] would say, at ADP, but we've made sure that Nadia's aware of not just our values, but our new leadership capabilities. When it's referencing... when the person's talking to them in their own language, Nadia will reference the leadership capabilities to reinforce those. We have a critical conversation model we wanted to make sure Nadia references. We integrated that pretty quickly. Nadia knows.

‍

We have a philosophy of being strength-based, so we wanted to make sure [00:58:30.366] Nadia was using strength-based coaching, which is, I think, what Nadia already does, but we wanted to make sure that the way that the interactions with the AI coach are going is in line with the way we actually train our HR people and our managers to be interacting with leaders.

‍

We were able to pretty easily, early on, integrate all of those pieces, and now we're doing things like, okay, make sure Nadia knows about our whole portfolio of leadership development programs, so if people are wanting some additional support, [00:59:00.579] Nadia can suggest certain things, or reference different practices that we have.

‍

Julia: It's a great question, because culture is so complex and nuanced. Can any tool fully understand an organizational culture? I don't know, it's a philosophical question, but I will say some practical things that we have done, similar to Rob, is we've trained it on our engaging leader framework, our leadership framework, so it does know [00:59:30.179] that.

‍

We also trained it on our individual development plans. We showed it, here's the framework, here's the worksheet, here are the things that we ask our employees to do. You can give it inputs, so it understands those things. From what I understand, you can give it as much or as little as you want. It's really up to you how much you train it on, but then it's only going to know those things.

‍

It's a great question, [01:00:00.519] like how much can you teach it so that it actually understands your culture? But yeah, the leadership framework and the things around individual development planning, I imagine performance management information for those who use it, would be the key things to include.

‍

Ellie: Did that answer? Fantastic. Yeah, Paul, fantastic question. I was going to go into the tech side too, so if you're curious about the tech side and where the context comes from, happy to talk about that later.

‍

Liz: If [01:00:30.019] you don't mind, I might add there.

‍

Ellie: Yeah, please.

‍

Liz: It's a process too. There were times where we looked at the data and we would find, oh, people are bringing up this thing, and we didn't realize they were going to bring up that thing, so let's feed that real quick, make sure it knows exactly what we want it to respond to. I don't know that I will feel comfortable until we've been a full lap around the sun to see all of the cycles that they're going to go through, all of the [01:01:00.196] HR processes that they're going to go through, and we can pick up on those trends.

‍

Scaling Coaching to the Frontline Workforce

How AI coaching reaches hourly and frontline associates

AI coaching extends developmental support to hourly and frontline workers who have historically lacked access to career coaches. At The Home Depot, where 300,000 to 350,000 of the 400,000-plus associates are hourly retail workers, Nadia provides on-demand coaching available across all shifts and time zones. Liz Ritterbush also described configuring guardrails so Nadia escalates sensitive issues to HR partners, relieving HR teams from being on call to memorize every standard operating procedure.

‍

Ellie: That actually leads perfectly into my next question, Liz, which is, as you're thinking about... You have such a wide variety of audiences getting coaching, that coaching looks different. Especially as you're thinking about associates at The Home Depot more broadly, what does coaching look like? How has that been implementing [01:01:30.159] that at the field and frontline level?

‍

Liz: One of the things I love about technology, and I think that there's so much potential with AI in particular, is around this coaching front. I was fortunate in that I had a career coach at one point in time. Not everyone does. Most people do not. In a company of half a million people, 300,000, 350,000 are hourly associates on a floor in a retail environment, they certainly don't have a career coach, [01:02:00.019] or access to a live coach.

‍

The scalability, something that's in their hands that they can go to: people are already using it. They're already pulling up Google, they're trying to make things act like a coach that aren't coaches. I love this opportunity to get something in their hands. It also supports... I mentioned that users were asking about HR practices. It supports our HR. When we've got people working Sundays [01:02:30.139] at 7 a.m., Saturdays at 8 p.m., we don't have an HR person that can respond to them right away. Having this coach that we've trained on all of our processes.

‍

We've also given it guardrails to say, "Hmm, red flag, I'm not going to give you an answer here. Go to your HR partner." That helps to relieve some of that burden on our HR who's trying to be that for our associates. It's too much pressure to ask our HR partners to [01:03:00.139] memorize every single SOP, every single line, and be on call 24/7. I love this idea of having something scalable that's for our associates for that.

‍

Ellie: Have there been any reactions that have surprised you from not just associates, but across The Home Depot?

‍

Liz: None besides my story, and I've heard of other stories or you mentioned other people sharing their stories about testing out the coach, "Oh, wow, we [01:03:30.119] can leverage it this way." Just breaking through that skepticism bubble has been... That was the biggest surprise for me. I do, of course, love the data that we're getting, and the trends, and being able to see, oh, people are actually talking about this thing that we didn't even realize was an issue, that maybe didn't come up in our online forums internally. Because here, they're talking to something that they trust.

‍

The Future of AI Coaching: Possibilities, Advice, and Open Challenges

What the next phase of AI coaching looks like

The next phase of AI coaching extends into 360-degree feedback at scale, performance review preparation with individual data inputs, neurodivergent-friendly customization, and broader democratization beyond the executive layer. Rob Lewis at ADP plans to use Nadia to deliver 360 results without training a large pool of HR coaches. Beth Shipman at Corning emphasized AI coaching's role in follow-through and behavior change, including reminders and nudges that human-led programs cannot easily sustain at scale.

‍

Ellie: Absolutely. I want to make sure we leave time for questions. Last question, as we're thinking about closing out today: as you think about your performance reviews, your processes, and these moments that we've been talking about, how have you seen the shift since you've implemented AI? Looking forward, what do you see as the possibilities in the future in [01:04:30.400] terms of support?

‍

Liz: I can talk to the possibilities. What I'd like to see: we know that users are leveraging it for performance management. We know that they're saying, "How do I have this difficult conversation?" I see leaders that are also very conscientious. They want to make sure they're putting in all of the right data, and they're giving it all the right information when they have that annual performance review. For us, there's two conversations, [01:05:00.219] but there's one big review. Am I including all of the right information? Can I remember everything that Ellie and I talked about? Can I remember every email that Rob sent and all of the impact that he had?

‍

There's an opportunity, at least for us, to say, "What are those data inputs? How are we helping you boil the ocean even further?" We're helping you boil the ocean when it comes to our SOPs. We're helping you boil the ocean when it comes to best practices and [01:05:30.059] leadership development practices, but are we feeding the AI the right individual data for you and the right individual data for your team?

‍

Knowing where people are tracking their performance data, and maybe someone up here has done this well, but it's something that I think is an opportunity, at least that I recognize. I also love the opportunities that it has for people like me who identify as neurospicy. We can program it. [01:06:00.010] We can say, "Hey Nadia, give me things in a bulleted list and give them to me step-by-step," and that's going to really help me. There's some opportunity there to have a coach that mirrors you and speaks your own language.

‍

Julia: In terms of future possibilities, love all those ideas. From our perspective, we tell people, and we really believe this, that General Mills is a great place to grow careers. Coaching [01:06:30.420] can be a part of that. It democratizes coaching. It's not just the top 1% of our employees, it's not just our director-pluses who have coaches. They do, and that's going to continue, but it's our other employees who are valuable, who we want to retain, we want to keep them, we want them to know you have tools and resources available to you. Please use them.

‍

It's [01:07:00.320] just communicating more and then seeing that flip around with more employees using it, talking to other employees about it. That's what I would love to see in the future, just having those casual coffee chats or water break chats where our hourly plant workers are talking about that. They probably already are in pockets, but I would love to see that become more mainstream.

‍

Ellie: Definitely. Rob?

‍

Rob: For us, [01:07:30.519] I think others have used Nadia for 360-degree feedback coaching, and we haven't done that yet. We're about to launch, actually, in the next couple of weeks, a new 360-degree feedback tool we've built ourselves on our new leadership model. It's a long survey, ordinarily, to scale a tool like this and give really quality feedback to each person who participates. You'd have to train a whole bunch of HR people or other folks across the organization to be schooled [01:08:00.309] in how to deliver feedback, how to help people interpret and understand, and make the use out of their results.

‍

Having Nadia there, where people can load up their results to the 360, and they have a coach there without all the additional training. By the way, not all your HR people are good at coaching or really like it or have the time for it. People sign up for it, they get trained, and they don't ever have time to deliver 360-degree feedback. This gives us the ability [01:08:30.140] to scale quality 360-degree feedback actioning in a way that we wouldn't be able to ordinarily. We're really excited about that as a prospect in the next few months.

‍

Beth: In terms of future outlook, I am excited generally. There's a lot of opportunity. A couple things that I was jotting down: certainly 360s, assessment, things like that. Oftentimes, [01:09:00.822] we want the best of the best, but that usually comes at a price point, and that also usually comes requiring someone to be certified in that exact instrument. I think this is breaking down the barrier to that. We would love to give everyone a human coach, and a person who just knows the best of the best and what it looks like. I think this helps every employee have that opportunity because coaching can be incredibly effective and accelerate [01:09:30.079] someone's own performance and their career path. That is part of our employee value prop, making sure we're giving those opportunities to every employee at every level.

‍

I love also the personalization. There's a lot more showing use cases. If you're the person who needs... Like, this conversation needs to be tough love, "Give me that feedback." I know my tendency is to operate in [a different way], but you can adjust that experience as [01:10:00.560] you go. Unlocking some of that.

‍

Looking at how we define ROI is another area I think we need to continue to figure out. It's not just doing things faster, but it's the effectiveness. If we think about growth and development, that takes time. I do fear that the accessibility is there, but I believe some of our stakeholders will go to the mindset [01:10:30.239] of training intervention, which is you took the training, you've gained the skill.

‍

If we know coaching, we know L&D. It's about follow-through, the behavior change, which takes time. It takes dedication, accountability, and I do love Nadia will follow up. You can build that in. She does nudges, she does reminders. The next time you're in a conversation, "Hey, how did that go? The last time we talked, it was about this topic. Give me an update." It's that continuation that I think otherwise [01:11:00.359] wouldn't naturally be there at scale. Opportunities left and right.

IO-to-IO advice for implementing AI coaching

The panelists' top advice for IO professionals implementing AI coaching: pursue specialized AI for behavior change rather than relying on general models, treat IOs as part of the design rather than passive observers, and start with small test-and-learn cycles. Liz Ritterbush emphasized that general models cannot be tailored for the kind of behavior change AI coaching delivers. Beth Shipman urged IOs to shape framework, terminology, and culture inputs. Julia Walsh recommended starting small, gathering feedback, and iterating change management.

‍

Ellie: I love that, absolutely. Last thing, and then we'd love questions from the audience too for our fantastic panelists: IO to IO, any advice for the audience as they're thinking about either AI coaching or the future there?

‍

Liz: I think see earlier conversation [01:11:30.600] on getting buy-in. A lot of people, there's this misconception, "Oh, I can train ChatGPT, I can train Gemini, I can train Copilot to do these things." What we're finding, and it's not just AI coaching, it's specialized AI. These bigger models, these more general models, you can't tailor them to get the behavior change that we're talking about.

‍

If you're looking for behavior change at scale, [01:12:00.559] focus on the specialized things like this and get the case studies. We found case studies were incredibly impactful with getting stakeholder buy-in, or like anything, get a good stakeholder to test it out. Get the engineers of the world to test it out and try to break it, so that you have some advocates on your side.

‍

Beth: I'll say, you're a part of the solution, I think. [01:12:30.100] We don't just sit back passively and say, "Well, this thing is going to take my job." You're part of the design, and you can shape it quite a bit, I would say, to be best practice, your framework, your terminology, your culture, how you can define it, your values. There's IO in that design.

‍

Julia: Yes to all these things. I would say test and learn. We started small, [01:13:00.420] and then we grew. We started small, we tested, we got feedback, we asked, what did you like? What did you not like? What was confusing? Then we used that to expand and grow, change our communication, change our change management. That was a really key piece for us as well.

‍

Ellie: Fantastic. Any questions from the audience?

What's keeping enterprise talent leaders up at night

When asked what challenges remain top of mind, panelists named cognitive overload, AI tool sprawl, workforce resilience, and the question of whether AI is helping or replacing genuine human work. Liz Ritterbush at The Home Depot pointed to behavior change versus information overload. Rob Lewis at ADP raised tool navigation across Copilot, Nadia, and skills marketplaces. Julia Walsh emphasized resilience and innovation. Beth Shipman framed the question as how AI changes the way colleagues relate to each other.

‍

Audience Member: What is something that's top of mind, or keeping you up at night in the talent and development space that maybe Nadia may help with? Maybe not, I'm just curious, what are things that are hard for you right now that you're thinking about solving?

‍

Liz: I touched on it earlier, but cognitive [01:14:00.684] overload, and how do we... And behavior change is skill, putting something in their pockets where they can act on it instead of the training that people forget that they took a month ago, or two months ago. That's where something like this could be really powerful, helping people boil the ocean, helping people reduce the cognitive overload from everything they're getting from Facebook or social media or from all of these other sources, [01:14:30.300] including work, is a challenge and something that we're working to overcome.

‍

Rob: For us, I think it's related to overload. We have not just the launch of Nadia, we have Copilot, we have a skills-based talent marketplace tool, we have function-specific technologies that are being pushed in sales, and service, and technology groups. We might have really [01:15:00.039] good tools, but are we articulating well enough for our associates when to use what? What tool works best for different use cases? How do we help people navigate that? The technology is getting better and better, but is it actually improving the experience of people, because they have too many things going on to know what to use for what? They might end up using a tool that is not perfect, but maybe that's okay. We have to help them understand [01:15:30.159] what the right boundaries and right use cases are.

‍

Ellie: Really quickly, Julia, before you go: we didn't get to this today, but Julia and General Mills have done a great job at the AI ecosystem and onboarding users, so [I'd] highly encourage everyone to go up and ask them about that after. Sorry, continue, Julia.

‍

Rob: Okay, go ahead.

‍

Julia: There's a lot of things keeping us up at night, but there's a lot of good too. I'll say just one of them, thinking about the rapid change [01:16:00.399] of the world and of skills and how do we make sure that our employees are feeling resilient through the change, that they're being brought along, that they're being a part of the change and not just it's happening to them, but they're changing with us? Building the resilience, leaning into innovation, and bold ideas. Thinking about that and how that [01:16:30.020] might not come naturally to everybody, but encouraging people to test and learn, try different things more, and get out of their comfort zone.

‍

Beth: Overall, how does all of this impact how we work with each other? That relationship, our organizational culture, trust. Is this workslop? Is this really how you think? Are you just reading what a GPT tells you to say? That's all how we look at each other [01:17:00.119] right now, and I think we have to figure out how much of that human-led, by design, versus AI-led makes sense for every company, and help our organizations through that culture change, and that's going to look different for everyone.

‍

Ellie: Thank you all so much. It's certainly one of my favorite parts of the job, working with all of you. Really appreciate your time and your expertise. If anyone has questions, feel free to come up after, but I think we [01:17:30.140] are at time, so thank you.

‍