Nick Caldwell on the engineering cultures that power Microsoft, Reddit, Looker & Twitter

[00:00:00] Engineers are often asked to go work on things without being explained to why they matter to the business. Just go build this feature and they're often asked to maintain or keep legacy code alive without understanding how. Every minute you spend explaining to an engineer of how it benefits a customer or how things ladder into the overall strategy.

[00:00:35] Welcome to in depth, a new show that surfaces tactical advice, founders and startup leaders need to grow their teams, their companies, and themselves. I'm Brett Berson, a partner at first. And we're a venture capital firm that helps startups, like notion, roadblocks, Uber, and square tackle company building firsts through over 400 interviews on the review.

We've shared standout company building. The kind that comes from those willing to skip the talking points and go deeper into not just what to do, but how to do it with our new podcast in depth, you can listen into these deeper conversations every single week. Learn more and subscribe today@firstround.com

for today's episode of in-depth. I am thrilled to be joined by Nick Caldwell. Nick is Twitter's VP of engineering, where he leads a 700 person team responsible for all of Twitter's consumer facing products. This latest role at Twitter is another step in an already storied career in engineering leadership.

He started his journey at Microsoft, where he worked for 15 years, rising up through the ranks to eventually become the general manager of power BI Microsoft's business intelligence. Next. He pivoted from enterprise and took on hyperscale by joining Reddit as its VP of engineering. He then returned to his BI roots by joining Looker as its chief product and engineering officer leading the org through its acquisition by.

Between Microsoft Reddit, Looker, and now Twitter, Nick's worked for companies with vastly different cultures. And in today's conversation, we comb through the biggest lessons from each of these orcs. With Microsoft, we unpack what Nick believes is a massively underrated approach to organizational design.

He explains the company's rigorous approach to regularly pruning and shaping the org chart. He also gives us an inside look at their management training and talent development, as well as what Nick calls the fairest performance review system he's seen. He also shares some advice for startups from Microsoft's approach to deeply understand and competitors.

As Nick tells it, there was a steep learning curve when he pivoted from 15 years at Microsoft. He doles out advice for others getting their bearings after a big career move. He also explains how Reddit's mission-driven culture informs his approach to leadership at Twitter. Finally, with Looker, Nick unpacks, his biggest lessons from leading both the product and engineering teams, which offered him a unique perspective on how these orgs that are often at odds can properly.

It's an incredibly wide ranging conversation. So there's something for pretty much everyone, whether you're interested in the cultural practices that power, some of the world's biggest companies, or you're a manager looking to level up, or you're an engineer with goals to take on leadership. Nick's got plenty of advice and insider stories.

I really hope you enjoy this episode. And now my conversation with Nick, thank you so much for joining us. I'm excited for our time today. 

[00:04:07] Yeah. Glad to be here. Thanks for the invite. I've been looking forward to it. 

[00:04:11] So one of the places that I think I thought that might be interesting to start and explore is having you reflect a little bit on the different cultures that you've worked in, because it's been an interesting journey starting with the first big chunk of your career at Microsoft.

And then read it and then look her into Google and now Twitter. And so I was curious when you think about each one of those chapters and you think about the culture and to be more precise when I'm using the word culture, I'm meaning like the way in which work is accomplished or what gets rewarded or all of the ways in which people work together to create some kind of outcome or the way in which people behave.

I'm curious what you've learned or observed because. Many different companies have different scales with very different cultures, all very successful in their own, 

[00:04:59] right? Yeah, for sure. And those are all radically different cultures. I think part of the joy I've had over the last couple of years of my career is getting to learn that there's a lot of different ways to ship software and a lot of different ways to lead and get things done.

I can run through some highlights from each of them. I didn't appreciate it at the time, because it's hard to understand the culture while you're in it. But Microsoft is phenomenal in terms of how it approaches organizational design and any other place I've worked relative to Microsoft has been less effective.

Forming teams, forming business units, understanding how to move people and resources to whatever the most urgent problems are on a regular cadence with some predictability tied toward a mission, vision, et cetera, that place in retrofit. Runs like a well-oiled machine it's clockwork in terms of how vision translates to strategy translates into org design, translates into people moving, et cetera.

And they do all of this while also maintaining a bunch of legacy or older products as well. And I appreciate that in retrospect, I think they're like the modern equivalent of like a GE in terms of. Just being good at organizational design management and so on and so forth. And it doesn't get talked about much publicly, but I think that is Microsoft's cultural superpower red.

It was the first consumer social software company I worked at. And that was where I sort of came to appreciate two things. And it maybe a bit surprising Reddit when I joined. Incredibly passionate about building a diverse team and had started to Institute a ton of mechanisms like hiring diverse leaders, changing how the hiring funnel worked so on and so forth.

And it was the first time I got exposed to accompany aware. If anything like the values came before, the product that they were making. If that, if that makes sense. And everything that we did was through the lens of does this ladder up to our values as a company, that the second aspect of that was just being very mission aligned.

The culture at Reddit was really around knowing that the site had been attracting deeply passionate users and moderators for more than a decade. And wanting to serve them and the employees at Reddit up until very recently weren't as concerned about business growth and so on. Like they, they were there because they just simply loved the mission of the company.

And then later on when Reddit started to expand and attract investor money and so on, so forth the monetary and growth objectives, intersected with the mission objectives. And I think they're off to the races now, but that was a real. Interesting deviation from Microsoft, which is ultimately like a big box, big business company.

And, you know, they, they're very competitive and red. It wasn't competitive in that way. Looker, for those of you who don't know, Looker's a BI company and the headquarters is in Santa. This has this introduced a bunch of like cultural oddities that I've not experienced anywhere. And I think it has to do with the fact that the headquarters was in such a remote location.

Santa Cruz is hard to get to, you have to drive through a two lane mountain paths to get there. So most of the employees live in Santa Cruz and Santa Cruz is characterized by. Kind of relaxed beach culture, if you will. Everyone knows everyone else. So with Looker, it's just going to sound very odd, I think, compared to how other companies talk about it.

But Looker is the first place I've worked, where I heard the CEO and exec team talk about the company being a fan. Every time I've ever heard that in any other context, I go, ah, full of shit doesn't mean anything. Looker is the first place I've worked, where they said it and they were genuine about it.

Like they genuinely cared about the employees in a way that felt familiar. And I saw activities from the CEO and the CFO there, which was. Just heart melting. I mean, like just the things that they did for the employees there went so far above and beyond what I've seen at any other corporation, like taking care of employees who got sick and, uh, looking out for each other in ways that went above and beyond.

I remember that too. I killed a bit of the cynic in me to see the leaders at that company and how they operated. Look, I got acquired by. And Google to all surprises that I, it was absolutely stunning to me as someone who never worked at Google. And if you're from Microsoft for a long time, you looked at Google as this super agile competitor.

At the time I joined Google really felt like a gigantic company. It really felt like it had all of the big company, innovation dilemma, bureaucracy that Microsoft had without. The skill of organizational design and so on and so forth. So I wasn't there for very long. And I think that Google is undergoing a reckoning on this right now.

I've, I've read a couple of news articles lately questioning their operating methods and so on and so forth. So I don't have a lot of detail on that, but it was shocked to me. I was expecting a culture. Characterized by innovation and what I saw it was not that, and it was a very surprise to me. And I've been at Twitter for a year, Twitter and Reddit have extremely similar mission-driven cultures with Twitter.

It's a little bit more on the nose though, because Twitter as a platform is. Something that surfaces all of the world's events, as soon as they happen, people go to Twitter to talk about what's happening in the world. And there's a lot of wild stuff happening in the world. And people at Twitter inherently understand that the power of this platform, that if you think about communications systems over the course of human history, from the printing press to radio, to TV, And so on, but each of these has allowed people, a new lever to reach a broader and broader audience and Twitter and social media are the latest incarnations of that and their orders magnitude more powerful than, than anything that has come before.

So everyone in the company sort of has this intrinsic understanding of the importance of the platform that we enable public communication in a way that is more powerful than any point in human history. And that has to be made available. As safely and responsibly as possible. And we see the impact of either success or failure toward that goal around the world in real time.

And this mission underlies everything that we talk about the company in a way that overcomes a lot of other faults that we might have on execution or problems. Like the mission binds us together. That 

[00:11:42] gives us some great points to jump off. So the couple things I wanted to chat about in a little bit more detail, I wonder to go back to your reflection of Microsoft and maybe have you describe what great organizational design looks like or stories that come to mind.

And one of the things that I think is just so. Under appreciated about Microsoft is the diversity of the businesses that operate under Microsoft. And they are just the world's best modern example of a multi-product company and a multi business line company. And I assume to your point. There's a lot of organizational design.

It's a lot about the type of people that they hire, but we'd love to hear you particularly given you were there for such an interesting, I think 15 year period. If you were to teach someone how it works or explain what the organizational design philosophy or how it shows up there, what that might look like.

[00:12:37] are some drawbacks to this too, if you want to cover them later, but a good parts of it. What the org design looks like is it starts with, as you mentioned, Uh, Microsoft is broken into multiple business units and they change these over time, depending on how they want to tackle the market or report to the street.

But you can think about windows, cloud, gaming, office, Ike, these sort of product families, and each of them is generally speaking. Multi-billion dollar internal business with even larger partner networks. So with that framing in mind, the company has to figure out a way to organize its resources. To address all of those business needs and the mechanics of how this works.

Are trying to say it, I'll try and simplify it within a particular business unit. You've got a silo or a VP who runs it and he'll have a, he, or she will have a corporate exact reporting into them and so on and so forth. But every business unit then breaks out, down into product families, and then within product families, product units, individual products, and then down from there into technology teams, feature teams.

Et cetera. And then in some cases there are support teams which provides shared technologies to all of the above. We're talking about within one product line, you get that entire tree of organization. What I'm describing here, you can look in the org chart at Microsoft and map back out what products are being used.

And there's a common saying, like, don't ship your org chart, but this is incorrect. Actually you should try it and ship your org chart. It's just, you need to have fluidity in your org chart to make sure that it maps whatever product or objective you're trying to achieve. Going back to what I said earlier, Microsoft on a very regular case.

Changes the org chart, like where the people sit, which products are owned by a particular product manager, et cetera, and communicates these changes very clearly, very succinctly like on a regular cadence. So you might expect that every quarter or every six weeks. That there will be some reorg announced by your executive and it will be backed up by, you know, Hey, we're doing this because we're going to go after this particular business strategy, or we need more people to, to shore up investment in this area.

I keep it at that level of detail, but it's a regular cadence. Of pruning and shaping the org chart to match whatever the product or goals that we're trying to get are secondarily, there are support teams. One challenge I see with many companies is that they neglect sustained engineering tasks or software fundamentals tasks.

Microsoft has an entire. I won't call them a product team, but there is an entire division which does nothing, but take care of internal tooling, software development, life cycle, et cetera. If there are products which are in sustained engineering mode, which has reached the maturity phase and will now no longer have any new feature development, but still need to be maintained.

Teams are formed to maintain those so that they're taken away from the responsibility of your main line product teams. This massively simplifies resource allocation in ways that don't quite exist at other companies. The final thing, which you may say as a pro or con, this entire structure and system is replicated for every major business unit.

Now. Pros of this are that the team working on gaming doesn't have to necessarily coordinate super tightly with the Bing team or the cloud search team. They can operate independently, run their own businesses. They have their own P and L et cetera. The downside of this, which is Microsoft is famous for, is that.

These teams might end up creating competing solutions to the same customer problem. And this is where I don't know if you've seen this funny diagram on the internet where they try and show a picture of the Microsoft org chart. And it's like every business unit has a gun pointed at the other business unit and.

Does happen. So like while Microsoft is very well organized to solve particular product lines, challenges that they can often come into conflict with each. So it can be a little bit territorial sometimes. And Steve bomber, who was the CEO for a long time, actually embraced this. You want it in multiple competing solutions with Satya, that's been less a part of the culture and he looks more for one Microsoft approach where all of the business units sort of aligned to overarching shared goals and reduces some of the conflicts.

[00:17:09] That was great. So how does zero to one product development work in the org design for 

[00:17:16] zero to one projects? The simple all answers doesn't work really well because Microsoft's a big company and any large company, Microsoft, Google, et cetera, is going to have a challenge with doing zero to one projects.

In general, there are two or three ways you might tackle them. Forming a V team or a spike team or pulling people from across the company into kind of a seed team and then isolating them and allowing them to focus on getting that new product that out the door is generally the approach that Microsoft uses.

In the past, they would do similar things through labs teams or entire organizations would sort of simulate it, startup incubators, but that pattern generally didn't work and as been abandoned by most companies, but a pattern that's frequent is to pull people together for a new initiative, spike on it and try and get to a V1 as quickly as possible.

The reasons. Tends to not work is in a big company. Anything that you might do in terms of making a new product could impact existing business lines. So if you're a PM at a big company, you can't just ship a new thing. I was on the power BI product at Microsoft, which was a zero to one BI tool, which is now quite successful.

But in the early days of it, when we were talking about building a new data visualization tool, We had to go and clear that idea with multiple teams like that would be impacted from office. The obvious ones are Excel, there's existing BI reporting tools, et cetera, but to even get moving on the concept, we had to clear a path to understand whether or not our product would cannibalize.

The preexisting business and you had to overcome a lot of inertia there. Even if you manage to do this, you have to sort of, it's not like a one and done, like you have to make time in your schedule or product roadmap to continually sort of carve out space for your zero to one. Until you manage to get to product market fit or some acceleration, but the best iterations of zero to one projects I've seen at Microsoft where pull the team together, spike as fast as you can, and then use middle managers and execs to provide air cover so that the folks who would claim you're cannibalizing, their business are held at bay until the team has enough time to get a fair shake at whatever they're trying to do.

[00:19:40] given that we're talking about Microsoft, I was interested. Are there other things that you've taken away? Or things that Microsoft did exceptionally well, that you've taken across your career. And I'm interested in the topic because I think that Microsoft is the most misunderstood or it's a company that operates in a very particular way that most companies in Silicon valley don't study.

Right? Like if you think about the way the Google operates. It's washed over almost every company in Silicon valley. And even over the past, call it five or 10 years, Amazon, I would say with some of their cultural practices, the way they operate an organizer more well understood. But Microsoft is a little bit of a black box and given the incredible success over such a long period of time, I was interested in any other practices or primitives that you find useful, or maybe founders today should be looking and studying and getting inspired.

[00:20:36] totally correct. And I've been in the bay area now, I guess for six years, I see what you're saying all the time, your Google management practices, the Amazon culture deck, all of that is really held on a pedestal. But my honest assessment of those things is that pale in comparison to the effectiveness of what Microsoft does.

And I have no idea why Microsoft isn't more open and about their practices. So there's. Maybe two other things I would say that they do extremely well. Microsoft scales through manager training. They have some of the most effective leadership development programs, the most effective leadership development programs I have ever seen in the industry.

And I've worked in several places and this comes from identifying top talent. They have what are called high potential programs. Other companies have this as well, but at Microsoft, they start very early on and there's a high potential program for every single career stage. So if you're an intern, a relative new employee, a senior employee, Staff level, there's a program for every single level to nudge you toward excellence at whatever that layer might be.

Additionally, Microsoft going back to org structure. They're one of the few companies which has career paths for ICS that go all the way up and they're clearly defined. And similarly they have career paths for managers, which lead into general men. So when you talk to, I don't want to pick on Twitter, but like every other company, every other company I've worked at has career paths that stop at being a great IC engineer or a really great engineering manager.

Microsoft's career paths lead into general management and then executive leadership. And there are book length analysis of it, like posted on their HR site. Learn quite a bit about management from the ground floor, all the way up to the highest levels, just by spending time on their learning sites. So long story short, their ability to develop managers as a superpower.

And to your question about what should people study? Microsoft doesn't publish these things publicly. I wish that they would just publish how they approach job ladders and management training and so on and so forth so that others could simulate it. But if you want to see a company that I think inspired this, I think it's actually GE I've read some Jack Welch's books and looked at how he developed leaders and how he structured his business units and so on and so forth.

And a lot of that, I would bet was picked up by gates and bomber in the early days to form Microsoft. And it seems to inform a lot of their management practices. Now, again, when you go do this, I'm not saying like replicate GE management tactics. Like these are very outdated right now. It just mean that the focus on management development and good organization performance systems.

I think there are seeds of all that content within GE that Microsoft picked up and has evolved over the years and just super briefly, cause I just reminded me of one other thing. Microsoft used to have a really atrocious stack ranked performance, the review system. Uh, 

[00:23:44] I was going to ask you about this.

[00:23:47] Yes. And I was going to say they got rid of that about I guess, seven or eight years ago and replaced it with what I've come to believe is the simplest most straightforward, fair, equitable performance review system. I have seen since, and I've tried to reproduce it in every place that I've worked and I just wished that they would publish what they do.

Simple easy to use fair tracks performance over time puts control levers into the right layers of the organization. So on and so forth. So I wish I had artifacts I could share about these things, but again, Microsoft is not right. Talk about its systemic approach to developing managers and performance in the way that other companies do.

[00:24:28] would. You just mentioned how impressed you were with that performance management system and how it's something that you've adopted. Could you explain the primitives there or how it works and maybe why you think it's a powerful way to do this type of. 

[00:24:40] So there's a couple of different elements to it.

First is how you assess yourself. Other companies use an OKR framework where it's very rigid. Did you hit this metric or so on and so forth? Microsoft has elements of that, but it's more introspective. It's more. Did you hit your personal goals? Did you hit your business goals and you fill out at a very high level, some goals alongside input from your manager that you want to achieve.

And you check on these things once a quarter, a basic starting point is this quarterly check-in, and then you may have metrics associated with those goals, but some of them are literally just personal development type tasks. You're encouraged to do those as well. Every so often you're going to have a performance review that is tied to.

Compensation as well. And here you need signal and input and that's how much of your goals did you achieve if you have metric-based goals or business related goals, what were the outcomes? How did they affect the business? Like all the things you would sort of expect because quarterly checking. Are happening.

You have a trajectory that's collected over time. So it's not just this, like every 12 months, you look at this long list of goals or OKR is, and that's a one point in time sort of thing. You can look backwards at the person's trajectory over however long they've been doing performance reviews and get an assessment of their momentum.

And this provides a lot of additional signal managers at a certain level, have the ability to have discretion over reward. So there are differentiated rewards. Like many companies have, but managers are given responsibility to spend budget toward their top performers and take away budget. If someone's performing at a lower level, all of this then gets reviewed.

Phenomenal oversight. So when you make your budget recommendations as a manager, they roll up to your director and then they roll up to your VP. HR overlays their checks on top of all of this. So that promotions are distributed equitably. We check that diverse employees are not being missed. So there's sort of a layer of oversight that goes into all of it.

And then additional. You can use this Florence review system too, because it's differentiated, uh, to signal when employees are performing really well or performing poorly signal that, and it all happens on a six month clock and you're constantly aware at all times. Of your top performers, your low performers, you're simultaneously rewarding people at a good rate.

And it's all done in a way that is, um, tracking equitable distribution at the same time. 

[00:27:20] What are the primary benefits of this system? Or maybe how you would compare it and contrast it to what might be a more traditional 

[00:27:28] model? I actually don't know what traditional isn't it. So I worked at five companies and they all had different that's fair.

Yeah. But the benefits of it are, it's a predictable case. You collect signal. If you get a new manager, for example, Without having clear history and consistent signal that new manager may have a different assessment of your performance or a different rubric that they're using with this system. You're over time collecting a history that includes your performance or feedback and your trajectory.

So it gives you a record that better informs your performance review at any given time, the predictable cadence of it, the fact that there's oversight, you can not run through this process and result in an outcome that right. Have an unfair inequitable distribution. Whereas in other companies we don't necessarily tie the oversight from HR into the performance review systems.

You run that risk. I like the fact that it starts by. Putting the onus on the individual employee to come up with personal goals that a track record gets developed and that it results in something that has equitable oversight. And it provides, again, signal to the management team about who the top and bottom performers are without being.

As onerous or obnoxious as a stack ranking system. So you're not operating in a culture of fear 

[00:28:51] in the stack, right? The main downside is the Lord of the flies competitive dynamic. That's right. 

[00:28:56] Yeah. Most of my career, well, early career before they switched was in stack rank meetings for performance reviews.

And those were very contentious and combative. You're essentially trying to go into those meetings. Not just to lift your team members up, but to find ways to pull other people down so that you can protect your top performers. And the dynamic of those meetings was never. Healthy. The intent, I think, was to find differentiated rewards and look for poor performers.

But the way in which it was done was frankly, unhealthy and combative and glad that they got rid of that system. 

[00:29:33] The threat, you mentioned just a few minutes ago about Microsoft manager training and clarity around managers. I was interested. Maybe you could share, how would you define a great manager as defined by 

It depends on the role that you're in, but at the highest level, a great manager is going to be able to define a strategy and execute against it effectively. And depending on what their discipline is, you know, engineering product, et cetera. By pulling together the necessary people into a strong team, coordinating with the cross-functionally dependent teams and then tracking toward clear progress.

So management at Microsoft is around coordinated execution. It's like a big ship, a lot of players on it. And you got to figure out how to get them all on the same page while marching toward a clear goal. And then related to that is. Doing so in a way that develops people's careers toward being excellent in whatever their function might have to be.

And then the final thing is that Microsoft is unique to other companies in the bay area. Is that if you're in management, if you're in a management role, You may start in engineering. You may start in product, uh, but you can lead your career toward general management in a way that allows you to have control over multiple disciplines.

In this episode with Nick Caldwell, we have a wide-reaching conversation on his storied engineering leadership career at Microsoft, Reddit, Looker, and now Twitter. 

Nick Caldwell on the engineering cultures that power Microsoft, Reddit, Looker & Twitter

Read the article inspired by this episode

“Get Off the Floor” and Other Career Advice from Microsoft, Looker, Reddit & Twitter