Katy George: Ditch the org chart - your team's future is fluid

This is not a once and done transformation where we all learn something, change the way we work, and then we're done and can go on to the next thing. No, this is going to be forever, that we are both doing our jobs and changing our jobs. Welcome to WorkLab, the podcast from Microsoft. I'm your host, Molly Wood.

Today we're joined by Katy George, corporate vice president of workforce transformation at Microsoft. We'll talk about where organizations get stuck moving from AI pilots to real transformation, what frontier firm leadership looks like in practice, and why the way most companies measure AI progress misses the point. Katie George thanks so much for joining me on WorkLab. Oh, it's a pleasure to be here.

So we heard your title which has workplace transformation in it, but I would love if you could just tell me a little bit about your job, because you've been doing this kind of work for a long time, and everybody's thinking about it right now. It's very relevant to know what that means. Yeah, we, to your point, we've all been talking about the future of work for a really long time, and now it's here for everyone. And I couldn't be more excited than to be part of Microsoft, really helping shape it.

So my role sits at the center of Microsoft in our Office of Strategy and Transformation and think of us as a learning accelerator, in that we have touchpoints across all of Microsoft, and we're looking for best practices and patterns about what really works to drive AI transformation, and how the roles of human beings are changing. And you just joined two years ago. That's right. So I would imagine that your job has already been transformed by AI.

What's it been like, you know, and what has changed, even in that relatively short period of time? my team does a ton of research. They do a lot of scenario analysis, workforce planning, models and all of that. You know, things that used to take months and we'd need an external contractor to help with, you know, whatever.

Now they're building on copilot and demoing, and it's just amazing how how fast, things have changed. And even I am not a technical person at all, but even I am now a builder, which is, really amazing. How do you think about. I just wonder, you know, just to sort of set the stage about your philosophy of workplace transformation.

Like, if you could boil it down to two or three things that have to exist within a culture, what might those be? Well, first of all, my strong belief is that when businesses change the way work happens in a way that really drives sustainable business results, which is really what AI is all about, it also makes the workplace better for human beings, and I truly believe that will be true for AI as well. And we together can help shape that to be true. And the second belief I have is that the only way you can effectively redesign work so that it works even better is by involving the people who actually do the work.

And that is probably more true in the context of AI, which is really all about knowledge work, where so much of what we do and knowledge work is tacit and it's not written down anywhere. So the people involved in doing the work have to be the ones who redesign it. And that's a super empowering point. Yeah, absolutely.

given that let's move into adoption, we're sort of at this point now where we have maturing technology as it changes every day, but it's maturing. It's available. Adoption is still struggling in some areas. what do you think of this?

What do we attribute this gap to right now? Lots of people either feel worried that I will eliminate their jobs, or actually make their jobs such that they don't, They aren’t good at them anymore. Like if you think about, our software engineers, you know, they have grown up as master craftspeople in coding. That's their professional identity and that's very powerful.

And now we're telling them you no longer have to code. Now you are a product builder. An end to end builder. And that in itself, when I you just think about, do I want to be a coder or a builder?

I'd rather be a builder. That sounds very exciting, but it is a real mindset shift. And so really bringing people along is is absolutely critical. Yeah.

And that is such a good point, though, when you think about the identity that you've built up in the the kind of personal equity you have in your skill. Yes. That transferring that over or thinking of it being replaced would be tricky. How do you bring people along?

I mean, I think that's we all sort of understand that implicitly. And yet it has to like, be structured. Yeah. So what we've we've been innovating, in a couple of different ways, around, you know, what we would traditionally talk about as change management, although this really is about setting up continual learning and adaptation skills, Right?

And mindset because this is not a once and done transformation where we all learn something, change the way we work, and then we're done and can go on to the next thing. No, this is going to be forever, that we are both doing our jobs and changing our jobs. So one of the programs that we run is called Camp Air, and that started in our engineering organization, and my team runs that, and it's now starting to spread to other functions. And what's special about Camp Air is that, team go as whole teams.

So this is not about each individual going off to a learning program when it fits in their schedule. It's about the whole team taking time out of their very busy schedule and and production kind of environment to, be in an immersive boot camp. The first week is all about talking about mindset, but also getting immersed in the tools. And we tell people, leave your discipline at the door.

So it doesn't matter if you're the product manager or if you're a software engineer or if you're a designer, everybody on the team is going to work together to get immersed in all of the tools that are relevant to the team's work, not just the ones that fit what you were doing in the past. And then by week two, they're starting to experiment and working differently and applying it to their own work, and then they're starting to apply it to their own work, and they still get coaching from other practitioners about how they can be most effective.

And it's really exciting. We do exit interviews with our engineers when they come out of this program, and they really have this huge shift in attitude. Even the ones who come in with the trepidation that I described come out as real evangelists, saying “I never want to work the other way again.” Because it is exciting to actually see, how all of the work fits together and really be part of the the total cycle.

You know, when you look at what's happening with software engineering, but we see the same in HR or finance or sales, is that sort of the cycles of getting work done are now so much faster. You're collaborating more in innovating and inventing and thinking about new ideas for for customer impact. And that is really fun. And it's, so I really do believe that the workplace is, is changing for the best.

what I really like about what you're saying is that there is a tendency, even on other episodes in the show, we have we have said to people, we have exhorted people, go learn these skills. Yeah, experiment, take time to write, try to figure out these tools and build an agent for yourself. That is ultimately individual and kind of puts the burden of keeping up with a changing workplace on the employee. How important do you think it is for companies to create things like Camp Air or structured experiences?

Well, what I think is special about Camp Air and some of the other programs we do is that we are not only thinking of AI as a set of individual productivity tools, what we recognize, and so the change management programs reflect this, These are about teams and organizational system change only if you change the whole system of work do you get different business performance. And so, whether it's a team based approach like that, where the whole team is changing to improve the outputs that they have, or we also run some other programs in other contexts.

For example, in our Salesforce that are role based, that is getting everybody who's playing the same role together, similar to Camp Air, is practitioner taught. This is not someone coming in from the outside is a grant facilitator. This is practitioner to practitioner. And these role based programs, one of them is a weekly huddle where literally everybody who is playing the same role in an organization gets together and they compare notes, not on what experiments did you do with AI in general?

But how have you used AI to get better at our job? And and what we're finding is that that's driving significantly greater adoption, but also much more effective adoption and effective adoption in the sense that we're starting to see the behavior change, that is actually linked to getting better performance. And it's because there's collaboration and there's encouragement. So, yeah, and and prior to this job at Microsoft, you, among other things, were at McKinsey.

Yes. Advising companies. And so I wonder, you know, when you turn that lens outward, what would you advise companies who might be struggling with some adoption? Well, first of all, realize that this is business transformation and this is not a product launch.

So, you know, we have this fabulous, collaboration with Harvard Business School around the frontier firm, research initiatives, etc. and they actually did a case example of our initial rollout of Copilot into our own Salesforce. And the first time we rolled it out, we thought of it as a typical product launch. Here's a new product.

We'll teach you the new product. And we expected, great things to happen. And people tried it and then they stopped using it. so only when we realized, no, this is about a business transformation.

We have to set clear business outcome goals. What are we trying to achieve? And then we need to really redesign work along with the people doing the work. By bringing together people who understand the tech, But also people who understand process.

So we do a lot of Kaizens, Gemba walks, value stream maps, day in the life analysis, etc. with people who really understand people, how are we changing all three of those things tech process and people together? All in service to clear business outcomes and success factors. And so, the first thing is to really understand that this is a business transformation you're driving.

And it has a lot in common with other business transformations that companies have led around other things in the past. But there are also some things that are pretty unique to AI. And this whole notion of trying to change knowledge work is actually pretty unique. I did lots of work in manufacturing, when I was at McKinsey and, you know, there when we were trying to change the way worked happened with Lean or then Industry 4.

0, at least you could see in the production line the widgets going down the line. Right? And you could see where they got delayed and there was inventory being built up, or you could see where there were deviations in the process that could cause quality problems in knowledge work. We don't see any of that.

so much is tacit. It's not standard, it's not written down anywhere. So, really understanding the workflow so that you can connect AI solutions into workflow and drive systematic change in the way work happens to accrue to business performance, it’s an art and a science. That's such a crucial point too because it not only gets your leadership and your managers on board, but it actually gives the people undergoing this training and adopting these tools a reason.

You know, a reason, totally. My team's now done 100 case studies of AI transformation inside Microsoft and looking at patterns and successfactors across those and, you know, one of the things we see consistently is this really wonderful combination of top down, bottom up. And what I mean is leadership is cascading down really clear business goals. You know, this is the focus.

These processes matter. This is what we're trying to achieve. And they are bringing together resources and creating space and time for people to learn and grow and experiment, but also resources to do some of the more complicated, you know, pro code, technology development to support AI transformation of those processes. But at the same time, they're also encouraging bottom up grassroots innovation.

And so some of my favorite case studies are grassroots, citizen developer, developed things that then are scaled to the whole organization. And so that combination of being clear about what the goals are, what success looks like, but then unleashing the creativity of the whole organization is really where magic happens. Yeah. How about on the, on the outcomes end on the other end of the transformation?

I know that businesses love metrics. Yes. And measuring outcomes. And that has been a tricky conversation in this space.

are the metrics for success? Well, it's interesting, I have to admit that when I came into this, I thought that was going to be relatively straightforward. And I think the world thought it was going to be straightforward because we all thought of AI as primarily an automation tool. So, you know, even most of the models that economists have used to predict what's going to happen with AI take job descriptions, break them down into all the tasks, figure out what percentage can be automated with AI, it's always 30% on average across any group of jobs.

You drop 30% to the bottom line. You reorganize your teams around the rest of the 70 and and you go and you have a clear ROI. That's not how it's happening. And actually, the real story from AI is far more interesting because yes, I absolutely is a great automation tool and that it can automate different pieces of people's jobs, not whole jobs, but, different tasks.

And some of that is quite dramatic, like the agentic coding that we're seeing now. But where AI adds even more value is something that we're starting to call “Capability Add,” where it's not just doing the same process with the same outcomes that you did before, but somehow just faster and better with AI automation, it's actually adding whole new capabilities in. So the value that's being created is actually a whole different level. I'll give you a really simple example that illustrates this.

Our internal audit team is awesome, and they have really leaned into really transforming themselves with AI. And yes, they have definitely found some productivity savings, some of which they have, you know, used so they didn't have to replace people with attrition. But, others, they've used to create better coverage for our company. So reducing our risk.

But what they would say is far more valuable than that, is that because of AI, every internal audit now delivers proactive risk identification. And that's not something that human beings used to do and now we've, you know, substituted AI for it. It's a new capability that AI has injected into the process. So one of the things that we're now increasingly trying to help our teams really understand is as you understand, you know, workflow, make it visible, and start thinking about where AI intervention would be most helpful.

Make sure you're asking the questions of “where should we be doing more decision analysis? More scenario work? Where should we have more proactive, or preventative kind of quality thinking? Where could we add value through this process that would delight the customer even more?

” Right? All of those kinds of questions. And therefore, it means that measuring ROI and the metrics gets harder, right? It was introducing entirely new variables to your business model.

Because now we're talking about growth, innovation, customer experience, employee experience, all of these things that are harder to measure and harder to create kind of a simple causation model that allows you to calculate ROI, but it's actually far more strategically important. You know, if you think about it, most companies do not compete on labor cost efficiency. maybe it would be, they'd get some benefit in the market if they were able to reduce their headcount. But actually to be really a sustainably exciting company with competitive advantage, the AI “Capability Add” that I described is far more valuable.

Yeah. what that means is that it is hard to measure the returns, we're trying to create more standards around some of those other metrics and how you actually think about accounting for and acknowledging, recognizing, kind of the growth and innovation benefits. We're also starting to experiment with a way of kind of mapping from adoption to metrics, which are the easiest thing to measure. Are people using the tool or not?

To behavior change, And then to the business outcomes. for example, in our Salesforce, we're doing some work where, we've used the, the role based change management huddles that I described, actually, to significantly increase, the usage and the effective usage of these fabulous AI tools that we have built with the people who do the work. And as we're seeing that adoption increase and the effectiveness increase, which we're looking at, are we seeing an increase in the amount of time that our sales folks are spending with customers?

Are we seeing them spend more of their time demoing and working on, you know, the latest and and most current part of our portfolio, right? That kind of behavior changes that we wanted AI to lead to. And then how does that lead to win rates in qualified pipeline? And how does that lead to what we really care about, which is revenue per rep?

And we see a really nice correlation between the reps who are using AI the most, and revenue per rep. But this is how we are kind of mapping kind of the connection between adoption of new tools to the ultimate business goal. What's so cool about this is that you have obviously been able to broaden the conversation about what these tools can do. And also as they mature, they themselves create new variables.

Yes. And ripples in the pond. Yes. That actually just makes everything so much more interesting and feels like you could say the companies like, look, if you just had a leap of faith about what you don't know, you could have and could be more creative about the way you get there, it's almost like you know, the Artemis II mission.

It was like, go there, look at the moon, tell us what you see. And you're sort of saying, deploy these tools and do a little bit of a science experiment, see what happens. Well, I'm saying that and I'm saying, spend some time thinking about, one of my colleagues calls it “possibility.” What is the possibility that I could provide to you?

And, Satya actually talks about this is what are your private evals? So if you're an engineer, you'll understand what that means. Like what really makes the secret sauce of your company? And what do you want to be accelerated there for with the AI system you put in place?

What is your unique competitive advantage that if you could accelerate it with AI would make all the difference? So identify that and then focus there. When I talk to customers, many of them say “boy, a year ago we were just doing lots and lots of pilots.” And now what we realize is that's a mistake.

We're now focusing on three end to end business processes that are much more boring, maybe, than the cool pilots who were coming up with, but actually will make the biggest difference in our business performance, and really drive competitive advantage. So that notion of being clear about what matters to the business and then focusing there and really thinking to your point about using an AI as the Artemis, kind of what's possible. and really exploring that is something that we think works really well.

So let's talk about this in context. You brought up the partnership Harvard Business School. And we are talking a lot about Frontier Firms. Yeah, kind of in the abstract.

But part of that partnership is to nail down what that really looks like, you know, to figure out the companies that have this mindset, tell us a little bit more about, you know, what these organizations do and feel like day to day. Yeah. Frontier Firm is where we are fully AI powered, right? Our decisions are informed by AI.

Our operations are sometimes done by AI, etc., but we remain very much human centered and human led. And one of my colleagues had a wonderful comment the other day that that when we're fully imbued, kind of with AI and the way that we all work and the way that our processes are working, it will feel the organization will feel more like a living organism and constantly learning, our organizations are learning every day because of the reinforcement learning that I naturally embeds into our organization, which I think is really exciting.

And so I think we will feel like the organization is much more organic, that we're constantly learning, that we're spending more time innovating, as humans and that hierarchy and job titles matter much less. You know, eventually, I think we'll finally get out of the, you know, 1900s, you know, matrixed org chart. And we'll be much more focused about coming together based on the skills we have to get, you know, really exciting projects done, that really leverage kind of our creativity and innovation.

So then what does that start to look like? As we also bring in or grow or train or recruit a new type of leader within those organizations, because you know that there are different management skills in a no longer 90’s matrixed organization. Exactly. I mean, one of the things that this means is, first of all, that I think leaders in a sense, have to become continuous experimentation officers, right?

They have to be the ones who are really encouraging, continual, adaptive kind of changes in their organization and teams. They have to be really focused on where value comes from. So I think that business understanding is going to be more important in more people than ever before. They have to be comfortable working kind of end to end.

One of the things we're seeing, even just at the individual job level, is that roles are becoming more T-shaped. Human expertise, still very important, judgment, very important, maybe even more important than before. That's something that, you know, I think a year ago, a lot of people thought maybe we will never need human expertise again, because we can all just access the world's expertise through AI that turns out to be false. In order to use AI really effectively, you need judgment, taste, expertise, but you're also then able to spend time kind of horizontally integrating across the whole process and therefore teams are becoming much more dynamic and fluid.

And so leading in that environment where you're constantly adapting to changes in the technology and in the system of work as it all evolves with technology, is going to be really important. this is really about creating an organization and a culture that thrives on continual adaptation that thinks that's fun and exciting, right? As opposed to draining. Right.

I think is really important. And then being really clear about being able to cascade kind of clear goals and outcome goals, down to the organization, because as we are all empowered, anyone in the organization to build and create, we need to know what we're building for. And that is going to become we're no longer cogs in a machine. You know, where we're each doing our job as it's been defined for us.

We now can be much more creative. But then we need to know what we're creating for. Right. I was just joking with a friend that, we didn’t think that there were any C-suite titles left, but I think maybe if it doesn’t already, Chief experimentation officer is a good one.

let's talk about, with this idea in mind that you have this kind of you have a different set of constituents as you are designing a workflow and a workforce for AI transformation, who needs to be in the room? Well, first of all, the business leadership has to sponsor and be actively engaged. This is not a tech project. This is not something that you know, your CIO leads on their own.

This really has to start from the business. In fact we were in a customer meeting and the CEO said “I am not leading AI transformation” and everyone went “Oh,” and he said, “I am leading a business transformation and AI is helping me.” And I think that's super important. So that's where it starts.

Even for example, in our customer support function, right? Which, so we're talking about call centers and things, right? The business has to be involved in every single part of the process of transformation. That's one of their key, key, principles for doing any additional AI enablement work, in this space.

So business leadership, sponsorship, and like, real active sponsorship, the leaders who are power users themselves are creating teams of power users. That is super clear. And then secondly, I think, you have to bring together a diverse set of cross-disciplinary skills to really change the system of work effectively. And so that does include, as I said, tech, process capability, like, you know, we're bringing with our continuous improvement teams, and people understanding.

And that has to come together. It can't be a handoff. And then the third thing I'd say, I would just reiterate, the people who are doing the work have to be there to be actually redesigning their work. And so those are the kind of ingredients we're finding for successful transformation.

I feel like you may end up repeating yourself a little bit, but I just want to do a little lightning round of kind of concrete advice for people. So a leader listening right now wants to move their organization from piloting to genuine transformation. Yeah. What are the first two or three things that they should do tomorrow morning?

Number one, get clear on what the most exciting, ambitious goal in your organization should be and therefore where to focus in terms of what processes and what success metrics. I just can't overstate how important that is to guide the efforts of your team. And then, you know, I'm really getting obsessed with the notion of cross generational teams working together on this, you know, early career folks bringing so much creativity and innovation, because they are, you know, AI native and then pairing them with people who really understand what great looks like and where value is created.

So, you know, think about diverse teams bringing the disciplines that I describe together. But also multi-generational. That's, is that happening here? That's a very cool point.

Yes. I mean, a lot of the examples that we see in fact work creating more, apprenticeship models, two way apprenticeship models. And we have a team, actually in our engineering function that's created a program called “Praise,” which is preceptorship for AI in software engineering. And it's literally pairing, you know, early career with more experienced folks.

That's very cool. This leads sort of nicely into the next question, which is as for the individual contributor who is using AI but maybe hasn't done all of this rethinking yet, or had some of this rethinking be communicated as goals, What is the first step to start to shift the mindset? first of all, you know, I can't emphasize enough kind of experimenting and playing around and enjoying it, right? Because it is really cool to see.

And if you tried something a few months ago, try it again, because the tools change so quickly, that you might be amazed at what you can do. Secondly, really talk to people about this all the time. Be obsessive. What are you doing?

What are you finding successful? And I think that's important both with other people in your same role, but also just other people. In fact, I was doing a, a video with a colleague, and we were both, you know, supposed to share examples of what we've done and she shared this example of something that I do all the time manually, like, oh my gosh, I never thought of doing that with AI, of course. Are you willing to tell us what that process was and have you changed it?

Oh, yes. So I am a total time management geek, so I really believe that, if you want to accomplish something, the first step is just making sure your calendar is aligned to it. Right? And I've gotten lots of time management coaching and it's really changed my life.

Anyway. So I often on a plane will literally get a notebook out, look at my calendar and, and in a really old school way, kind of think about, okay, how have I spent my time in the last month? And does it align to the things that I think are really most important? And she just asks copilot to do it.

Duh! So, yes. I haven’t, this just happened. I haven't done it yet, but I absolutely am going to do that.

I'm going to want you to report back, that’s amazing. Absolutely. I love the idea that to learn AI better used to talk to more humans. Oh, by the way, also talk to AI because, actually, you know, Copilot has some pretty good answers if you want to find, you know, tips and tricks for how to use it even more effectively.

Very true. Katy George, thank you so much for the time today. I really appreciate it. Oh my pleasure.

Thank you for having me. Thank you again to Katy George, corporate vice president of workforce transformation at Microsoft. For more conversations from the frontier of AI at work, follow the show and listen to past episodes wherever you get your podcasts. I'm Molly Wood.

Thanks for listening.