There are numerous forks in the road along an engineerâs career path. Do you want to stay an IC, diving deep into the codebase and shipping new features? Do you set your sights on management, climbing the ranks and eventually leading a large org? Do you stay on board as a small startup hits scale and gets bigger and more complex, or does the pull of building from scratch draw you back to an early-stage environment? And finally, do you have an eye for entrepreneurship, with your own goals to become a founder?
Over the course of her eight-year stint at Stripe, Amber Feng faced each of these questions.
As a new grad, she was one of the famed startupâs first dozen hires and climbed its ranks from scrappy IC, to first-time engineering manager, to starting her own teams from the ground up and directing hundreds of folks under her orgs.
Itâs quite atypical for engineers to spend eight years at a single company â particularly with countless job opportunities piling up in their inbox. So plenty of folks quiz Feng on why she bucked the trend and stayed put for so long. âWith Stripeâs trajectory, the job was changing all the time. Over the course of my eight years there, it felt like eight different companies. Too often folks hop to the next company opportunity prematurely, leaving a ton of untapped learning potential at their current company and role,â she says.
When Feng eventually did start to feel the pull to leave, it was to forge her own path. She teamed up with Mahima Chawla and Lauren Dai (a fellow Stripe alum) to found Cocoon â an employee leave management platform (of which First Round is a proud seed supporter).
In this exclusive interview, she guides other engineers up and down the org chart as they weigh their next career moves, leaning on the frameworks and lessons sheâs pulled from in her own career. First, she uncovers a few under-the-radar traits that set the most impactful engineers apart from the pack. Next, for folks weighing the decision to move into management, she gets real about the four pieces of advice she wished she knew when she first became an EM. Finally, she turns her attention to the past couple of years building Cocoon as a first-time founder â particularly the elements of Stripeâs operating principles that she's weaving into Cocoonâs cultural fabric. Letâs dive in.
THE UNDERRATED TRAITS OF THE BEST ENGINEERS
âWhen people talk about great engineers, thereâs a lot of emphasis on writing code, designing systems and building products. All of that is extremely important. But I see lots of folks who are drawn to engineering for engineeringâs sake. Theyâre hyper-focused on the type of system they want to build, regardless of whether it actually solves a problem, to the detriment of their long-term growth,â says Feng.
The best engineers Iâve worked with arenât necessarily the ones who write the most complex architecture, theyâre the ones who can think through a problem really deeply and are endlessly curious.
She sketches out a few traits that tie together the standout engineers sheâs worked with over the course of her career.
Theyâre crisp writers and communicators.
At Stripe, the company enforced a strong writing culture, which Feng has since brought along as co-founder of Cocoon. But in the early days, it wasnât a skill that came as naturally to Feng. To shore up her own writing skills, she kept a close eye on the strongest folks at Stripe and noticed a few key areas that the best communicators had in common.
- Concise. âPeople often think being a strong writer is about using flowery language. But one of the biggest steps is actually figuring out how to be concise. How can you convey the most important information, like updates, blockers, highlights, risks, etc. with just a short summary?â
- Tailored. âItâs about learning how to read the room and understanding your audience. Whether itâs talking to non-technical folks or talking to customers, you need to be able to put on different hats and adjust your communication accordingly. Whatâs the essence of what this audience needs to know?â
- Anticipate and address the FAQs. âWhether youâre writing a doc, an email or an announcement, youâve got to anticipate what questions might come up. You can even ask folks around you what their own questions are about your project. When you force people to go through that thought exercise, they start to put themselves in the shoes of whoever is reading the email or the customer update.â
Feng also flags another underrated aspect of great communication that deserves your attention. âIâve found that at the heart of great writing is bringing others along and helping them understand the decisions that youâve made. If youâre part of an early team, sure youâve built the foundation of the codebase and the architecture. But, more importantly, youâve built the culture and the quality bar and the decision-making frameworks, and these need to be captured in writing so they can live on across the org,â says Feng.
Great communication is about more than just saying what youâre trying to accomplish. Itâs about helping folks understand how you make decisions so they can be able to make similar types of choices in the future.
They aim for the broadest impact.
âWhen we talk about engineers, too often we hyper-focus only on whether you can ship code quickly or design the most scalable or robust system. But the more important question is whether you can drive actual impact at the company,â says Feng. âAs a founder now, I often say that Iâm hiring entrepreneurial generalists â people who think deeply about the problem theyâre solving, not just from a technical perspective â maybe itâs an organizational problem to solve. They constantly question whether they are even building the right thing.â
Her advice for engineers who hope to take on an ownership mentality? Try to get in the room where it happens. âItâs really important to be able to see other people drive projects end-to-end and examine what made them especially sharp. Think about how you can get yourself in the right roles or companies where you feel like youâre learning on the ground,â says Feng.
And thereâs no replacement for getting as close to the customer as possible. âYou can hear anecdotally from user research or look at a product brief that explains the problem, but there is no substitute for jumping on sales calls and listening to what customers have to say,â says Feng.
Engineering doesnât operate in a vacuum. Who are you actually building for? Reduce the feedback loop as much as possible. Literally, go find your user, sit next to them and see what theyâre doing. Become steeped in the problem.
âWatch them use the product, see what buttons theyâre clicking and which ones theyâre not. What are they confused about? What flows do they use?â she says. At Stripe, the company reinforced this habit from the beginning by adding engineers to the support rotation, with engineers getting paged if a customer question hadnât gotten a response.
âBeing able to chat synchronously also helps. Email is slow â add them on Slack, IRC, Discord, text message, whatever makes sense for your company,â says Feng. And as much as possible, try out the actual product from a userâs POV. âDogfood your own product to understand what the problem is, not just what the feature requests are. Youâre not trying to build a faster horse, youâre building the car.â

CHARTING YOUR ENGINEERING CAREER PATH: IC, MANAGER, OR FOUNDER?
Before becoming a founder, Feng spent eight years at Stripe â but plotting the points on her time with the company, you wouldnât see a traditional pattern. Rather than climb a traditional ladder, sheâs gone from IC, to manager, to exec, back to IC many times over. Drawing from that well of experience, she doles out four lessons for first-time EMs.
Know you can get off the track.
âWhen I first joined Stripe, I was a bright-eyed and bushy-tailed new grad. At the time Stripe was just about ten employees and there werenât any real teams. But a couple of years later, I was leading our product engineering org that had grown to 30 people. In an alternate universe, I could have continued climbing that ladder,â she says.
Instead, she took an alternative path. âI kicked off a team called Treasury, which was basically responsible for tracking all of Stripeâs money movement. After starting that from scratch, a couple of years later that team had grown to 100 engineers,â she says. Once again, she went back to building 0 to 1 within Stripe, starting a small team that built Stripeâs corporate card and banking products.
So her advice to other engineers who are weighing whether to stay an IC and build or to go the management route is that the paths can overlap. âYou can swing back and forth â itâs what I did, and itâs what Iâve seen a lot of amazing engineers do. Listen to that gut feeling. If youâre in a role you think youâre in love with because of the title or how many reports you have, you probably wonât be incredible at it,â says Feng. âThatâs not to say that any role will be 100% things you love doing all the time, but itâs important to balance the energy-draining things with the energy-giving things.â
When it comes to career aspirations, itâs important to differentiate your altitude on the org chart from what you actually enjoy doing day-to-day.
âOne of the most important things to remember about being a manager is that your impact is no longer just your own â itâs the team that youâre managing. If youâre someone who gets a lot of energy from coding or building systems, you may not enjoy having an indirect relationship to those problems anymore,â she says.
Encourage, but donât shield.
âThe biggest lesson for me is the importance of being brutally honest with yourself and your team. I had a misguided sense that it was the managerâs job to be overly optimistic or positive, even when something wasnât going well,â says Feng.
âEarly on at Stripe, we were working on a big project to rewrite a huge part of our product. It was one of those classic projects that was destined to fail because youâre chasing a moving target. If I look back, it didnât feel good in the moment and folks in the room knew it wasnât going well. But I felt the need to stay really upbeat and encouraging,â she says. âAs you can imagine, it ended in the worst way possible. We abruptly killed the project after a year of sinking precious time into it.â
Many well-intentioned, first-time managers think their job is to sidestep uncomfortable truths.
âIn retrospect, I should have been having way more conversations about the fact that it wasnât going well. I think a lot of first-time managers are well-intentioned in this way. They want to be supportive. They want to make sure people are happy. They think that honestly talking about the risks will be demotivating. But itâs often the opposite â people want to rally around problems,â says Feng.
Hereâs how sheâs tactically made that shift. âAsk a peer or a friend to keep you honest. Every week, have them ask you three extremely direct, uncomfortable questions about your project. âWhy is the project delayed? How do you know whether the team achieved the right outcome? Is a person on your team underperforming?ââ says Feng. âIn the hectic day-to-day, itâs easy to justify slippery timelines or let these important questions slide to the backburner. You need someone to pull you out and who you trust to keep you accountable.â
âIf you have a spidey gut sense, acknowledge it. Are you feeling nervous about a project? Why? Whatâs wrong with it? Others probably have that sense too and itâll feel much better to be able to talk about it and problem-solve together,â she says.
Next, create a safe space for the team to share their own concerns. âAt Stripe, we often reflected on one question: If Stripe failed at the end of the year, what was the most likely culprit?â Do a hypothetical exercise with your team of projecting outward â if your project or team failed, what was the most likely reason? Invite your team to share their FUD (fear, uncertainty, doubts) in a safe space. Youâd be surprised what critical context or ideas folks offer up in these honest moments,â says Feng.
Give feedback like an Olympic coach.
Feng also points to a sneaky undertow when it comes to feedback in tech. âWhen youâre a high performer, the extent of the feedback you receive is, âYouâre doing great!â And while that makes you feel good in the moment, itâs not actually helpful in the long term,â she says.
âI think about if I was an Olympic gymnast or an NBA basketball player and I asked my coach how I was doing. If all they told me was, âYouâre doing great!â Iâd probably fire them. Your entire job as an elite coach is to give brutally honest feedback. I want to be world-class at what I do, so I want you to tell me what I can do to improve, even incrementally.â
We want to be world-class Olympic athletes in what we do â and we need brutally honest feedback to get there.
When it comes to doling out impactful feedback for high performers, Feng recommends starting with the reportâs career goals. If you donât know what people on your team care to learn about, youâll struggle to give them useful feedback unless theyâre actively underperforming. âIf I have a star engineer on my team who wants to become a founder, all of a sudden I have so many ideas for them. The feedback probably wonât only be about improving their code quality. The feedback will be about what they can do to take on more leadership with one of our big business lines, to get them closer to their ultimate goal,â she says.
Operate with scale in mind.
âWhen youâre an IC and youâre hands-on-keyboard coding, the feedback loops are typically quite fast. Youâre seeing the impact almost immediately. But as a manager, thereâs a mindset shift,â says Feng. âWhen I say Iâm a roll-up-my-sleeves kind of leader, I donât mean that Iâm trying to get involved in every single thing that happens. I mean that I like to deeply understand the things that are happening and try to provide thoughtful advice or frameworks, so that when someone comes to me with a thorny problem, we can have a meaningful discussion.â
Sheâs instituted a new ritual to put that philosophy into practice. âMy new rule for myself is that when someone asks me a question or asks me to make a judgment call, I try to write out my thought process, rather than just a direct answer. I want my team to understand how I came to that decision so they can make those calls in the future for themselves,â she says.
Advice for future founders.
For engineers with ambitions to become a technical founder someday, Feng reflects on her own transition â returning back to what gave her energy. âThe part that I was really excited about was all of the different pieces of company building: assembling the right team, figuring out what our culture and operating principles should be â not just building a product or a system,â she says.
And although she did plenty of 0 to 1 building at Stripe, adjusting to the stakes as a founder was a steeper incline. âIf youâre part of a bigger company working on a product, you have a lot more resources to lean on. You can cry for help and get more support. But as a founder of your own company, the buck really stops with you. You have investors or peer founders to lean on, but you do have to get comfortable with that existential shift,â she says.
For folks weighing whether or not theyâd like to jump on a founder opportunity, she sketches out a few must-haves:
- Mission. âYou have to be incredibly excited about the mission â not just the product youâre building. Are you committed to tackling the problem? In the early days, the product can change so much. You donât want to get too caught up in excitement about a particular solution or technical challenge.â
- Team. âFind the right co-founders who are complementary to you. My superpower is in execution, whereas Iâm weaker in the higher-level future vision setting. I care about being internally facing rather than an external spokesperson. It was important that I found co-founders who balanced these with their own strengths.â
- Forecast. âMy co-founders Mahima, Lauren and I went through the Co-Founder Dating Questionnaire, which gives you a preview of what it would actually be like to work together. I highly recommend any potential founding team go through these questions. Some cover how youâd split equity or if you want to IPO versus get acquired. Other questions are more philosophical around values alignment. You donât have to agree on everything â in fact, itâs probably better that you donât. You want to see how folks can engage in healthy debates about different perspectives.â
BUILDING A NEW COMPANY CULTURE: BORROWED LESSONS
Although Stripe changed drastically in the eight years since Feng had joined the company, the reality she was facing when co-founding Cocoon was that she had really only been exposed to one company culture. Her task, like any first-time founder, was to pluck out and borrow the bits that she wanted to weave into her new companyâs cultural fabric. She explains the ways in which Stripeâs culture has influenced Cocoonâs â along with one other tactic the company borrowed from Apple.
Honing product instincts across the org â not just the product team.
Stripe notoriously took the unconventional route in the early days of not having any product managers. âThe thinking early on was that we were engineers building an API for other engineers. We wanted the feedback loop to be as fast as possible. We would be g-chatting our users (before Slack existed) asking what they thought of the product â getting digitally as close to our customers as possible,â she says.
But once the company began to expand beyond the exclusive developer audience, Stripe brought on more product folks. âAs the product gets more complicated and there are a lot more moving pieces, you need that organizational glue from someone with an end-to-end perspective,â says Feng.
While early on this approach certainly worked out for Stripe, too often, Feng sees this lesson that you donât need product managers applied too broadly across the tech scene. âStripe had a really strong foundation of engineers with stellar product instincts. Itâs less about whether you have specific people with the product manager title, and more about looking at the skill-sets at the company. Are there people who happen to be engineers or founders that are doing the job of a product manager already?â
While Cocoon has product folks as part of the early team makeup, diverging from the Stripe model, Feng and her founding team have layered on that product lens across the org. âOne of our foundational values is that everyone works on product â not just people whose titles have the word âproductâ in them,â she says.
This goes beyond the typical engineering, product and design trifecta. âWeâre building an infrastructure product thatâs very technically driven, but at the same time, itâs an educational product. People are learning about how to go on leave for the first time. All of those touchpoints, from onboarding to helping companies craft the right leave policy to giving advice to a first-time parent, are all about the product,â she says.
Appoint a conductor.
But every person at the company weighing in on product decisions would lead to a chaotic symphony of voices without a conductor. âFirst, the founders need to set a North Star that clearly articulates what the company is aiming for, and how youâre going to make certain trade-offs. For example, one of our big pillars for this year is how can we scale to 10x volume? As an operationally-heavy company, we want to make sure that weâre scaling with technology, not scaling with people,â she says.
To avoid too many cooks in the kitchen, while maintaining that product ownership culture across the company, Cocoon borrows Appleâs DRI framework. âThe notion of the Directly Responsible Individual is that we want everyone to have ownership for the company and the product that weâre building, but that doesnât mean that everyone is a blocking decision-maker. The DRI is the ultimate person who is accountable for the goal. Itâs their responsibility to make sure theyâre looping in the right stakeholders, making sure people feel heard, and actually making the call when there are conflicting opinions,â says Feng.
Hereâs how it works in practice: âWe have four pillars for the year, and we have one person assigned as the DRI for each of those four pillars. Theyâre responsible for figuring out how we execute and assembling the right team to tackle those problems,â she says.
The key here is to remember that DRI is a fluid role, not a title. âIt doesnât need to be a specific engineer or product person, and someone who is a DRI for a pillar one year might not be a DRI another year. You need to appoint someone who is able to step back and think about the overall goals and outcomes, rather than requiring someone in a specific title or seniority level,â she says.
Debate without ego.
Feng also points out another sprinkle of Stripe on Cocoonâs cake. âThereâs a lot of talk in founder circles about hiring nice people, but I donât think itâs just about hiring folks that are nice. You want people on your team who are kind and generous, but also who arenât afraid to push back on something they donât think is quite right or debate passionately for the thing that they believe in,â she says.
After growing up in the Stripe org, where passionate debates were commonplace, itâs a cultural touchstone Feng is constantly reinforcing at Cocoon. âPatrick Collison built this relentless culture of questioning what we were doing and whether it was the highest impact thing,â she says. âAt Cocoon, we talk internally about debating without ego. Itâs not about being prideful or defending something just because itâs your idea. But we want folks to speak their mind without being afraid of being seen as ânot nice,ââ she says.
And if youâre struggling to come to any sort of consensus, try taking it back to square one. âSometimes you need to narrow down what youâre actually disagreeing on. Maybe one of us is missing context or information, or weâve misunderstood what the goal was supposed to be. Itâs about maintaining mutual respect and high trust for one another, even when you disagree,â says Feng.
This article is a lightly-edited summary of Amber Fengâs appearance on our podcast, âIn Depth.â If you havenât listened to our show yet, be sure to check it out here.