In this edition of Author Talks, McKinsey Global Publishing’s Jessica Marshall chats with Dean Carignan, who works at Microsoft’s Office of the Chief Scientist, about his new book, The Insider’s Guide to Innovation at Microsoft (Post Hill Press, Winter 2025), coauthored with colleague JoAnn Garbin. Through seven case studies at Microsoft, Carignan and Garbin uncover common patterns among high-performing, innovative teams. They explore the role of culture in innovation and share how other companies could benefit from operationalizing their approach to transformation.
An edited version of the conversation follows. You can watch the full video at the end of this page.
Why did you and JoAnn Garbin write this book?
My coauthor, JoAnn Garbin, and I wrote the book as part of our mission to democratize innovation. We saw how fast the world is changing and how, with generative AI, it’s going to change even faster. We realized that every person and every organization is in an innovator’s dilemma.
There needed to be a handbook for how to navigate those situations. We also realized that in our employer, Microsoft, we had the perfect reference material. It’s a company that has innovated for 50 years. It has reinvented itself fully four times.
We decided we would study Microsoft, distill what was general and applicable to any person in any company in any industry, and bring the book out to the world on the eve of Microsoft’s 50th anniversary. We also made a commitment to donate all proceeds to STEM education charities with the goal of cultivating the next generation of innovators.
You mentioned the innovator’s dilemma. What is that?
At its core, the innovator’s dilemma is a situation in which the very things that have made an organization strong in the past make it weak in the future. We tend to repeat the things that have helped us to get where we are.
If you’re not willing to take a hard look at what you do and whether it’s really adding value to others, you won’t be able to identify an innovator’s dilemma. And you won’t be able to disrupt yourself.
What about Microsoft’s approach to innovation made you feel that there was a story to tell everyone?
As we started conducting the research, the most interesting insight was that nobody started by talking about the technology that they were innovating.
They always started talking about people, processes, and culture—things that are universal to any organization. As that pattern played out over and over again, we saw that these are not just practices of innovation that are from Microsoft, or are just for the technology industry, but practices anyone in any organization can use.
The book delves into seven case studies, and the first one that you write about is Xbox, Microsoft’s gaming platform. What did you learn about innovation from interviews with the Xbox team?
I love the Xbox case study because it allowed us to delve deeply into the role of culture in innovation. Everyone talks about the importance of culture, but Xbox has very specific practices that it puts in place to guide its culture.
There are three practices that we really zeroed in on in the research:
- Rehiring the team. Any time the business model makes a radical pivot, Xbox makes a dedicated effort to revisit every employee’s scope and focus, helps them make the transition, and ensures they want to make the transition. It rehires the team.
- Managing the culture. The second investment the brand has made is implementing a dedicated, centralized team whose job is to monitor the culture and help it evolve. That team manages culture in the same way that you might manage a product. They have metrics, reviews, and processes; they make investments in culture. They take oversight very seriously.
- Recruiting. The team recruits not for culture fit but for culture add. They really want to bring people into the organization who help the culture evolve in the direction that it needs to go and who move it toward the places that customers and partners want it to be.
You mention that the Xbox team told leadership, “We really need to do this differently than you think we should.” Why was that important?
We open the book with a story from Xbox. It’s the story in which Robbie Bach [former president of entertainment and devices at Microsoft] and a handful of his colleagues went in front of Bill Gates and Steve Ballmer [former Microsoft CEO] and said, “We’ve got to be in the gaming business. It’s going to involve building a piece of hardware, a console. It’s not going to run Windows because Windows is for productivity apps. And the name Microsoft will not appear anywhere on it because that’s not what gamers are looking for.”
They were dismissed from the executive offices based on that vision, but they kept bringing it back. And eventually, Bill and Steve said, “OK, we get it—you’re funded. Go and create this new line of business.” We love telling that story because so much of innovation is about persistence, drive, and believing in your vision and being able to go back time and time again until you convince the people who need to be bought in to go for it.
So much of innovation is about persistence, drive, and believing in your vision and being able to go back time and time again until you convince the people who need to be bought in to go for it.
Some might be surprised that Bing, Microsoft’s search engine, is one of the innovation case studies in your book because it’s not the dominant tool for search today. How did Bing leverage the ‘underdog advantage’ at Microsoft?
The Bing case study is one of my favorites in the book. It’s so rich because it goes way back in our corporate history to the point when we had won what was called the “browser wars.” We won by being a fast follower. We became complacent, and we missed the search revolution. We got into the game with Bing, but we were too late.
Fundamentally, Google innovated not only the technology of search but also the business model. They created a flywheel where users brought advertisers, revenue, and better algorithms. They built an unbreakable competitive model. But Microsoft didn’t give up. We acknowledged that being a market leader would be challenging but that it was very important to be in search.
Bing was given the mandate to compete with Google, to continually take a little bit of market share away in small increments, but to do it with a fraction of the budget. Here’s what we love about that point—and we see it again and again: Constraints breed creativity.
That’s what we call the “underdog advantage” in the case of Bing. My absolute favorite anecdote from the chapter on Bing describes how its machine learning team transitioned to a new form of machine learning—deep learning—way before the industry did.
They did it partially because they saw the potential but largely because they didn’t have enough people trained in the more classical approaches to machine learning. Deep learning, of course, is at the heart of what we now call generative AI. The other thing that Bing did was to realize that competing against Google was hard. That made Bing very good at certain functions, such as indexing the web, categorizing large corpora of data and information, and machine learning.
Bing became a hub, a center of excellence, which would share those capabilities across the company. A lot of what users experience using Office, Azure, or our other products originated with Bing and was then shared and scaled across the entire company.
Therefore, taking that head-to-head competition with Google and turning it into a hub of innovation for the rest of the company was just the perfect eruption. It was also one of the ways that Bing has helped us to transform as we transition to AI.
AI is a big piece of Microsoft’s innovation story, including how to grapple with innovations that might cause harm. How can companies think about innovating responsibly?
When I look at Microsoft’s journey in terms of responsible AI and responsible innovation, I commend our chief scientific officer, Eric Horvitz. He spent his entire career focused on human-centered AI, which brings into consideration many safety and responsibility concerns.
In 2016, he anticipated that AI would be a big, important wave. He set up a small organization, of which I was a part, called Aether, which stands for AI ethics and effects in research and engineering. We started to ask: Where is AI going? What are the risks? How can we prepare?
That led to policy efforts to write responsible AI policies, engineering efforts to build tools and safeguards, as well as training efforts to make sure all our employees understood the risks that could come with AI. We wanted to ensure they were working that into their day-to-day efforts to manage and mitigate risks.
The fundamental challenge with responsible AI and innovation, more broadly, is the pace of change that we’re experiencing. With AI specifically, every time the models improve, the risk surface expands.
When we completed our original evaluation on GPT-4, relative to GPT-3, I remember how much more capable that model was, which meant it could do many more positive things but also many more harmful things. Every time a new model or a new technology comes out, you have to reassess: What are the risks? Who are the users who could be affected?
The fundamental challenge with responsible AI and innovation, more broadly, is the pace of change that we’re experiencing. With AI specifically, every time the models improve, the risk surface expands.
The way we’ve worked to address that at Microsoft is to create this incredibly big and multidisciplinary tent of people who are constantly looking at the new models and systems, assessing the harm, and building tools to mitigate that harm.
That tent includes every discipline in the company. We’ve got policy people, we’ve got user experience people, engineers, and scientists. That allows us to do quick sprints. When a new technology arrives, we can quickly evaluate it, look at the different kinds of harm it might cause, reflect them in our policy, put in place compliance motions to make sure we’re tracking them, build systems to reduce the risk and to measure the harm, and do that almost on a weekly basis. Since the landscape is changing so fast, we have to change with it. The secret is a multidisciplinary and agile approach because AI isn’t standing still for us.
You close each chapter with separate insights by you and your coauthor. What perspectives did each of you bring to the process?
JoAnn and I were the perfect pairing to write this book because I am a longtime “intrepreneur,” 20 years at Microsoft, and JoAnn is also a longtime entrepreneur, but most of her career has been in start-ups and midcaps. She really knows that world.
When we finished a chapter, we would sit down independently and write about what most resonated with us and how we might’ve dealt with it ourselves. What was incredibly provocative about that process was that whatever we would write independently, there was a version of it for the corporate world and a version for the start-up world.
It was that process of writing our notes that really gave us conviction that the practices are universal. Even with our very different backgrounds, our diagnoses and prescriptions were incredibly similar.
What are some ways that Microsoft has operationalized innovation, and why is this so important?
When we write about operationalizing innovation, we’re really trying to dispel the myth that innovation can’t be planned or managed, that it’s just serendipity—an idea falls out of the air, and we get lucky. Every group we studied had a defined process and way of working through innovation.
It’s so critical to systematize it because every time you innovate, you’re asking people to bet on you. You’re asking finance to invest. You’re asking other teams to pivot the ways they work. You have to be systematic and convey that credibility to get people in line and working with you. And systematization doesn’t necessarily mean it’s predictable or as linear as the way the rest of the company might operate.
Operationalizing innovation means going through three phases: discover, design, and develop. Being able to communicate those phases—where you are, where you need to go, and who’s going to help you in each phase—brings a sense of structure, order, and logic that helps other stakeholders in the company to buy in and to find the role that they can play to help.
It’s so critical to systematize it because every time you innovate, you’re asking people to bet on you. You’re asking finance to invest. You’re asking other teams to pivot the ways they work.
What surprised you the most about researching and writing this book?
The biggest surprise was how joyful the innovative organizations were. We even coined a term for it: “serious joy.” They were deadly serious about what they were doing. They worked hard, hustled, and were rigorous and disciplined.
Yet they were having so much fun. We realized that’s a critical part of a healthy organization. Innovation taps into our human need to create. When you have that environment, irrespective of what people are building or how, where, or when they’re building it, there’s a joy that you often don’t see in more routine parts of the company.
We were not seeking serious joy. We didn’t have a hypothesis about it. In every interview and in every discussion, it came through. We realized that we ourselves are joyful when we’re innovating. It shouldn’t have been a surprise, but it was.
Can any company apply these approaches to their innovation?
JoAnn and I deeply believe that the patterns and practices in the book apply to anyone in any industry, not just technology. Part of the reason we wrote case studies was to allow people to go inside Microsoft and see the blow-by-blow descriptions of the innovative process.
When you do that, you realize that it has so little to do with technology and everything to do with processes, culture, people, and systems. We recommend a couple of things to people who want to start on this journey. First, think very deeply about your distance from your customer. All innovation is grounded in your knowledge of customer needs. If you’re not bringing through that customer signal, you’re not going to innovate in the right direction. The second recommendation is to look hard at your culture.
Every organization says, “We want to be more innovative; ideas are the lifeblood of our organization.” But you have to ask yourself whether you have a culture that encourages risk-taking, that delivers psychological safety, where people feel like they can take chances, that trains people and gives them the tools for innovation so they can approach it in a systematic, structured way. Looking both outside and inside your company and asking those hard questions are probably the most important first steps you can take to become more innovative.
Watch the full interview
Dean Carignan shares how you can create a culture of innovation
