Facilitating the Spread of Knowledge and Innovation in Professional Software Development

Write for InfoQ


Choose your language

InfoQ Homepage Presentations Mission, Culture, and Values: Using Them to Guide Your Company through Good and Challenging Times

Mission, Culture, and Values: Using Them to Guide Your Company through Good and Challenging Times



Heather McKelvey discusses LinkedIn’s guidelines used to weather events, such as economic downturns, and how to turn those periods into opportunities.


Heather McKelvey is VP of Engineering @LinkedIn.

About the conference

Software is changing the world. QCon empowers software development by facilitating the spread of knowledge and innovation in the developer community. A practitioner-driven conference, QCon is designed for technical team leads, architects, engineering directors, and project managers who influence innovation in their teams.


McKelvey: You're sitting here wondering, why is somebody going to talk to me about mission, culture, and values at a technical conference? Because it's going to guide a lot of what you do with your career. It's going to guide a lot of how you choose companies. It's going to guide a lot of what you're going to do if you've found a startup. It will help you through both good times and bad. We're going to talk a little bit about me. We're going to talk about how those things play into caring for your career, how you can nurture not only yourself, but others. Then we're going to talk about how it can help you through both challenging times and opportunities. Culture, mission, values are key for us to be consistent in making sure that we're giving ourselves and others the right guidance.


I have been a software engineer for 42 years. I started out writing protocol drivers for TCP and SNA in assembler. Trust me, those drivers are not around anymore, or those companies are probably out of business. I am VPE for services infrastructure at LinkedIn. What we do is we build the infrastructure that the rest of the LinkedIn engineers use to build the LinkedIn platform and applications. This is things like standardized access to data, services that run in production that auto-tune JVMs. Things like traffic infrastructure. At LinkedIn, we do the CD pipeline for them as well. LinkedIn really does do continuous deployment. In that, if an application engineer gets to ship it from their tech lead, they can be through the whole CD pipeline, through the canary, through the tests into production with 1% of the traffic on them, within 30 minutes. Pretty interesting infrastructure. I have been an engineering leader at six startups. To my credit, four had nice exits, two went under, but we can't all get them right. What I found is, over the years, the mission of the company, the vision they have, their culture matters. Wasn't too long into my career, when I figured out, this is what I value in mission. This is what I value in culture. If my expectations do not match the company's expectations, I'm either trying to manage me out or they're trying to manage me out. It's a good thing upfront to get that stuff right.

Couple of quotes to start us off. I think this is really important. Nobody's going to remember the deliverable you did last year. Nobody remembers the protocol drivers I wrote, except me. What people will remember is how you made them feel. Did you help them transform? Did you support the mission and the vision? How we do that to a certain extent is through our values. That's the contract upon which we agree to work together and build trust. I love this quote from Maya Angelou, and I think it's something we all need to take with us. Second quote I've got is from Jeff Weiner, and I think this is really important. Your mission should be pretty consistent. Your vision should support your mission. It may need to be fine-tuned over time. Your culture needs to support transformation. It needs to support transformation of individuals, teams, and business. I think that's an important one as well.

Culture and Values (Values Re-enforce Culture)

I put LinkedIn's culture up here, because it was easy for me and I could copy it. Many companies have their own mission, vision, and culture. These are the values by which we focus on. This is the contract by which all individuals at LinkedIn are expected to support. Word of warning, if the CEO does not believe and uphold the culture and the values, you're trying to manage them out, or they're trying to manage you out, I'll guarantee it. Three things I want to point out here that are really important, trust and care for each other. Trust and care for each other, when things are good. When things are going really well, we celebrate each other. We celebrate our wins. We celebrate how well the company is doing. When things are not going as well as we would like, or, I made a mistake, I need to be able to apologize, we still trust and care for each other. We give each other that important feedback, so that, one, as individuals, we can help each other transform, grow. We can help the team transform and grow. Then we can also tell senior leadership, you need to transform and grow. This is an important point. If we don't trust and care about each other, we are not giving feedback. I've had a situation more than once, where someone has come to me and said, I'm being challenged with a senior leader, whether it be my manager, or an IC who's a tech lead or whatever. Thought I'd ask, have you told them? Have you told them how this is affecting you? Have you told them how this makes you feel? No, I can't say that to them, so you don't trust and care for them. You don't have their back. We are open and honest and constructive. This goes very much with trust and care. The other part of this, to round it out, it's the belonging, it's the diversity, it's the inclusion. When we talk about inclusion, we're not just talking about gender, we're not just talking about ethnicity, we're talking about people who think differently. People whose abilities are different from ours. We're talking about people who may have grown up in an environment where they didn't have the same economic advantage as you. Making sure we're being as inclusive as possible, is very important. Giving those people the same opportunities is very important.

We took this culture, and for the services infrastructure team, we decided we're going to use some of this as some of the basic foundations of our guidelines for our charter. If you've never written a charter, I strongly suggest you write a charter. You should start with three dream big ideas. If you only have one, great, but don't go more than three because it's way too much. We said, how are we going to use these? Since we are an infrastructure team, we said we're going to use these to put our customers first so they can put our members first. We did this because we're infrastructure, when we get it wrong, 7000 engineers are screwed. Not a good position to be in. We've affected their productivity. That is part of what we're doing there. We said we're going to renew and modernize the stack, and automate upgrades as much as possible. Our CTO says to me, "Renew and modernize the stack? Everybody does that. No." We get into a habit sometimes where we are revisiting something that was maybe decided upon 10 years ago, so we decided to build it ourselves. Because, yes, I know they build it over here, and there's this open source thing over here, but I'd like to build it myself. Not necessarily the best idea, because sometimes the industry will catch up and overtake. An active open source community is extremely powerful. You want to be part of that community. You want to be able to contribute, give back. You want to be able to help the rest of that community make an even better open source product.

Then the last thing we said is we're going to automate resilience, remediation, scale, and deployments. Why did we do this? How many people here can tune a JVM? It's what I thought, less than 10. We wrote a very interesting library that runs in all of our instances, that detects when an instance is sick. There are three triggers to figure out if an instance is sick. The most obvious one is, I can't get any CPU utilization, n number of times over so many milliseconds, so I'm going to start load shedding. I won't even give the request to the application. The first time we saw a whole cluster of over 500 start load shedding, somebody the week before had tuned the JVM by hand. That was emphasis enough for us to write services to make sure that we were automatically tuning the JVM for them in production.

Career Culture and Care, and Learnings

These values, these are the contracts we use. You start using these pretty much early on in life is my finding. In my career, I've gone through what I consider three stages. The first stage is, find my passion. You're early on, you're trying new things. What's going to make me happy to get up and go to work every day? It's not usually unique to company, it's something that I want to pursue for my career. Finding your passion is extremely important. It's also giving yourself the bandwidth to try different things, finding out what you don't like, and saying, I know why I don't like that. Absorbing as much as you can. This is a great time to learn quickly, learn fast. These values are important because when you start to use these values, you understand you may be early in your career, but you're there to help the whole team. Then you move to your second stage. In the second stage for me it was, hone my expertise. I found in my early stage, I really liked distributed systems, they're hard. They're complex. They need to be able to degrade gracefully. They need to have pieces be able to fail and no one perceived it failed, except those of us who got the monitoring alerts and went, uh-uh. Distributed systems was extremely important to me. I also found at this stage in my career, the values that I was using for guiding myself and others through our culture became important to me to guide myself. I found, using them to establish my boundaries was key for me to be able to be more successful over time.

One of my favorite stories is I was staring at this stupid bug for two days, and I could not figure it out. My husband kept having these talks with me about, "You need to set down because you need to be home at a certain time. I want to see you dear." I'm like, "You're in marketing, I'm a software engineer." I know your day ends at 4:00. I'm staring at this stupid bug for two days. I get this call, and I was pleasantly surprised to hear my husband say, "Woman, shut your laptop." I just started laughing my head off. I was like, "I'm coming home, it's 8:00, I know," and get in the car. Fifteen minutes later into the drive, because I'd walked away from it, I deal with the problems. Setting your boundaries, making sure that you are taking care of yourself, making sure you are getting what you need. Remember these boundaries will change over time. In your career, you will at times have a huge appetite for growth. Then that will go down, things will change. Something may happen with personal life, or just be, you just went through a major growth spurt and now you're like, "I need to settle." That's fine. Use the values to set your boundaries. Pay attention to the flow of your life. I remember one time I was a new manager and a female software engineer, she was a senior software engineer. I was like, "Come on, let's talk about getting you to staff." She looked at me, and she said, "I have two small kids, I'm not going to staff." I'm like, "Never mind." Three years later, she was ready, once they were in school.

Use the set part of your career to also remember, you are now in a phase where it is more important for you to teach, educate. Use those code reviews to leave a lasting artifact of teaching for others to discover. Start with seeking to understand. Remember we talked about inclusiveness. People don't think the same way. Some people's approach is very different. They may all come to the same conclusion, or some of them because they're thinking very differently, may come to a very different conclusion. Seek to understand people's approaches. As well as once you can start that discussion, you now discover, how are they comfortable learning. That's an important thing to do. You're going to have a greater impact on people and their transformations. Then this last stage, it's not about me. I'm not the expert. I learn more from my team and other teams than I could possibly know. It gives me a chance to go, I can go back to playing like I did early in my career. I can go back to learning as much as I want. I can actually bring clarity to what we're trying to solve, because I'm asking questions, because I don't understand. Always be building and learning and paying it forward. Because at this stage of your career, when you get to really strong leaders, it's not about their transformation. It's about how they helped others to transform. It's about how they help teams to transform. In the end, that transforms the business.

Weathering Tech Industry Ups and Downs, and Learnings

These values, why is this culture value thing so important? The tech industry isn't infallible. In 42 years, I've been through at least tech downturns, three of them. Probably more. I've been in the business so long, it's ok for me to forget some of them. First one was in 2003, the dot-com bust. This was really the tech industry. In 2003, some of you will remember and some of you will be like, "I don't know what she's talking about. I was born in 2004." There was an approach to startups where you had to start with $5 million, $10 million. It was based on the valuation. That's how the VCs figured out how they were going to value your company. What was your most recent valuation? Over 100 tech companies actually didn't make it through that bust. The VCs learned during this time that they needed to be much more pragmatic. They needed to change how they value companies. They need to look at the founders more closely. They need to look at how things are being managed more closely. They started measuring startups by burn rate. The founders started realizing, when I go ask for money, and I need money, I don't get money as easily than when I don't need money. They started learning from that as well.

Ok, go forward, 2007, 2009. Most of you have lived through that. Maybe not as closely as some of us did. Global Financial Crisis. This wasn't just tech. They were closing down banks. There were bank takeovers. I worked for a company, a startup, and Sequoia was our major investor. Our CEO had been CEO for three or four months, he was a brand-new CEO. Sequoia called in all the CEOs of the startups, when the Global Financial Crisis kicked off. If you haven't seen it, Google it, it's their infamous Rest in Peace Good Times. They made it very clear, money wasn't coming for at least two years. This was a pretty big financial crisis. Many people got laid off in many industries across the globe. What came out of this one was interesting. Engineers vote with their feet, they find something new to do. They're innovative. They're creative. If I can't find a job, I'm just going to go do a startup. That happened a lot. In the Bay Area alone, in 2010, there were 280-plus startups founded in one year. Angel investment went through the roof. I don't need that $5 million to 10 million anymore, I can just do this, just give me $600,000 to start with. That'll get me and my two founders off the road for like a year and a half, and we can build something. Then the other interesting thing was, I can't afford a data center, and I don't want to buy a server. I actually don't even know if I want an office, but I can very cheaply spin this up in AWS Cloud. People started architecting for the cloud, not the thinking that, "I've got this in my data center, I'll just move it to the cloud." When you start architecting for the cloud, you are building distributed systems. You are building something very interesting. You are building something that allows you to scale with less boundaries. AWS revenue through this whole thing did really well, it helped them take off immensely.

We get to now, 2023 state of tech. VC money has dropped again. Over 170,000 tech workers were laid off according to Crunchbase, between 2022 and 2023. The most important part is we're going to see these ups and downs. We may have to pivot. We may have to refocus our team. We may have to, unfortunately, sometimes shrink. We may have to take advantage of it and grow. The culture and the values should be the same, and they should help guide you through these difficult times. It helps you make those difficult decisions. It helps in having the conversations about the difficult options, and where should we take this. One other thing, we talked about inclusion. I just read, Harvard did a study for 85 years. That's longer than I've been in this business by two. The study was about, what is really most important? Is it purpose, or is it money? Turned out, consistently, purpose was far more important than money, regardless of the age, regardless of the gender, regardless of the person. Purpose allowed people to, one, better choose what they want to work on, where they want to work, who they want to work with. Make sure that they have the same culture, they believe in the same mission. It ended up that purpose had a huge benefit for people, in the fact that they were happier, they were more satisfied. It was an even better reason to get up and work on that mission every day. If you're doing a job interview, and if you're me, and you're like, tell me about the culture. Then you ask the question 15 different ways because you're doing the lawyer thing. You're trying to catch somebody up and hear a different answer. Ask them about what purpose do they want to have in their lives. What do they think? This should lead directly back to that mission. It will lead back to the culture. That's an important point, because when those things match, the impact you, the team, the business can have, is much bigger than any of us think. Why does it take 85 years to do that study? I'm not sure.


Just to round this out, keys to success during challenging times. You realize that these are grounding rules for you not just living your life, but how am I going to make the best out of my career? How am I going to feel I am actually impacting not just me, not just the team, not just the business, but hopefully others as well. This is an important one, I think this is really good. We've had people go back and forth on this, which is, dream big. If I dream big, I might not be able to achieve it. If you don't dream big and you can't achieve it, did you just limit yourself? Did you just limit the team? What if you dreamt big and you only got 50% of the way there? Still pretty good. When you get 50% of the way there, you might see the path to get to the other 50% of the way there. Don't be afraid to dream big. Don't be afraid to make those statements. Culture reinforces how we work together. Remember I talked about that third phase of my career. It is not unusual for people in the first and second phase to think, the CTO is an expert at everything. Truth is, the CTO is an expert at nothing. They have so much they have to do. You have to enter into discussion, seek to understand what they actually know, what they are seeking to learn, to help them and to educate them. When somebody in a leadership position asks you a question, it should not be, I have the answer for you. It should be, can you set that context for me? I'm trying to understand better what you're trying to work through to get an answer to. Then, think about it. Spend some time, write down where you want to make an impact. Write down your purpose. Think about your values. Use them. Be clear about them with people. Don't hide them. Use them to evaluate your next opportunities, because you'll end up like me, when they don't match, I'm trying to manage me out, or they're trying to manage me out.


See more presentations with transcripts


Recorded at:

May 29, 2024