How to Build a Great Engineering Team
Over the past few years, I’ve had the opportunity to build out a world-class engineering team at Paxos, a fintech startup on an ambitious mission to enable the movement of any asset at any time in a trustworthy way. As VP of Engineering, I had to figure out how to build a strong team as we were assembling the rocketship. Considering how crucial our technology is to our mission, it was actually like refueling the rocketship mid-flight while it was being assembled!
Here are some of the most important things I learned along the way. This list is neither comprehensive nor prescriptive, but rather, a collection of things that I have learned from my time at Paxos.
Hire Great People
This goes without saying, but hiring great people is easier said than done. The market for great engineers is extremely competitive, and startups usually don’t have a recognizable brand or the deep pockets to spend on marketing to candidates as large tech companies do.
Who to hire? First things first: A-players attract other A-players. Finding the first few A-players for a startup is the hardest step. While this chase can get frustrating, you need to be patient. It’s important to not give in to the urge to fill a position with someone who’s not quite an A-player, which can be tempting at first. If it’s hard for you as the hiring manager to say “no,” then make sure you have a bar-raiser in your interview loop. A bar-raiser is a skilled evaluator of talent within the company who plays a crucial role in the hiring process. The bar-raiser should be somebody outside your team who won’t give into the temptation of making a hire just because there may be an urgency to fill a role. Bar-raisers can be those who are not even in the same area of expertise as the interviewer.
The best candidates come from your own network, especially previous colleagues you have experience working with. These individuals hit the ground running, and you won’t need to dedicate time building rapport with them. Unfortunately, most companies have a non-compete clause in their employment agreements, making it difficult to recruit individuals from your previous employer. Other potential strong candidates within your network include classmates or people you may have hit off instantly at conferences or meetups. Sourcing from your network means frequent coffee chats, drinks and events for you to tell your company’s story.
How to recruit? As VP of Engineering at Paxos, recruiting great people is one of my primary responsibilities. To do an effective job, you need to get good at storytelling. You should be able to tell a compelling story about why they should work at your company, why they should work in your team and why they should work on your product. People often talk about selling your company’s vision to the candidates, I like to think of it the other way round. Candidates should be able to buy into your story. Work on your story, write it down, use visuals and refine it based on results.
How to interview? After sourcing candidates, design an interview process that creates a great candidate experience. It’s important to design outcome driven interview questions to ensure that interviewers look for a signal. At startups, while there is a tendency to have the entire team interview a candidate, this doesn’t scale beyond three engineers- it can be overwhelming for candidates to meet so many people.
The interview questions should be battle-tested for outcomes that are predictive of performance on the job. I often see interviews riddled with puzzles and questions designed to trick the interviewee. Not only are these absolutely the wrong kind of questions, but they also create a bad interviewing experience. Your interviews should focus on similar types of problems that you would come across in your day to day work. Not only does this create better engagement with the candidate, but it also helps with evaluating candidate responses. You already know your systems and have spent time thinking about them far more than the candidate, so you’ll be able to see the difference between a great answer and a not-so-great answer. The other important thing to assess for is the core values of your company. Developing questions that test for core values will help ensure the candidate is a good culture fit, which can’t be overlooked.
How to calibrate across interviewers? To ensure a consistent, strong hiring bar, it is important to calibrate across interviewers. Start by creating interview guides that are constantly updated based on candidate feedback, interviewer experience and candidate performance data. By creating interviewer onboarding guides, new interviewers will quickly learn how to evaluate candidates. I highly recommend pairing a new interviewer with an experienced interviewer to shadow and learn from them. Once the new interviewer is ready to interview, the experienced interviewer should reverse shadow the new interviewer. This ensures that the experienced interviewer can step in if needed to steer an interview in the right direction. After the interview, the experienced interviewer should provide feedback to the new interviewer.
How to make hiring decisions? Every interviewer should be looking for some kind of signal. For example, one interviewer may be evaluating for problem solving skills while the other one may be evaluating for design and architecture skills. Interviewers should provide feedback in the company’s applicant tracking system (ATS) right after the interview. This is the most important step. Simply noting “Hire/no hire” is not good enough. Interviewers need to log detailed feedback, which can be especially important for the hiring manager to use when there is a mix of “hires” and “no hires.” Detailed feedback allows the hiring manager to make educated tradeoffs depending on the role they are hiring for.
How to continually enhance the entire hiring process, from sourcing to offer acceptances? Be diligent on collecting hiring metrics! Metrics will allow you to identify bottlenecks in your hiring process. You should be able to answer the following questions:
- How many people do you need to source to get 1 hire?
- How many offers do you need to get 1 hire?
- Where are candidates dropping off in your process?
- What can you do to improve your conversion rate?
- What are your top of the funnel stats? Is your sourcing engine working efficiently?
Constantly look at data, learn from the data and tweak your process to see if you are getting better results.
How to grow the team? Hire senior engineering leaders proactively. People always fall into the trap of hiring engineers first and leadership second when it becomes unmanageable. This is almost always wrong in a high growth startup. Hiring leaders early on is incredibly helpful since they will help the company grow exponentially. Leaders will each have their own teams to grow, allowing for the even distribution of recruiting responsibilities amongst more leaders, leading to more hiring throughput.
Build an Engineering Brand
A startup is constantly competing for talent with big tech brand names like Google, Facebook and Amazon. One of the ways you can start to build a brand is to start an engineering blog. Top engineering talent is always excited to hear about interesting problems that your company is working on. A-players want to work with other A-players. So, get your best engineers to blog about interesting aspects of what they are working on. From my experience, blogs are definitely worth the time investment and are great for recruiting.
Presenting at meetups and conferences is another effective brand-builder. In New York City, where Paxos is based, we have a thriving blockchain and cryptocurrency community. There are a lot of meetups where our engineers have been able to showcase their work. We have consistently been able to get our name out there and source good engineers at these meetups. Conferences also help us network with great talent. Establishing thought leadership by presenting at conferences and participating in tech panels increases the visibility of your engineering talent.
Contribute to open source! If you’re a tech company, chances are you’re using open source software. Don’t just use it, contribute back. OSS contributions help improve your engineer’s brand and the company’s brand. For example, Paxos engineers have contributed to popular open source projects like Bitcoin, Pycoin, Lightning Network and Hashicorp Vault. We have also open sourced some of our libraries for other engineers to use. These contributions not only help build your networks but also solidify your position as a company that gives back to the community.
Build an Impact-Focused Culture: Outcomes over Outputs
Outputs are what you produce or deliver, but outputs may or may not result in meaningful or desired change, such as a change in customer behavior or change in product engagement. Outcomes are a more appropriate measure of meaningful results or the value we create. Deploying Jenkins is an output. Deploying Continuous Integration/Continuous Delivery pipeline that improves our deployment velocity by 150% is a quantifiable outcome. You may decide to deploy Jenkins to satisfy the outcome, but that’s not the goal. Outputs are “what” we produce where as Outcomes address “why” we do something. Build a culture of quantifying the business outcomes to create a sense of purpose in the work produced.
Always prioritize outcomes over outputs in your decision making process. It’s not uncommon to see a lot of opinions when there are smart people in a room. All those opinions probably sound great, but at a startup you’re always short of time and resources. You need to ruthlessly prioritize what you need to work on. The best way to do this is to focus on the “why.” Encourage people to present their opinions, but be sure to explain the “why” behind those opinions. The “whys” will help you understand what is important for the customer and prioritize the work. At Paxos, we use techniques such as Hypothesis Driven Development and 5 whys to great effect. It’s extremely important that you build these mechanisms and embed them in your culture. They will pay off significantly as the organization grows to democratize the decision-making process.
To help engineers delve into the “why,” make business impact the key driver behind any engineering-related decision point. Engineers love to be at the forefront of technology and gravitate towards bleeding-edge gizmos to solve the next problem. While learning new tools and keeping up-to-date is important for engineers, it’s also important that they deliver value and impact to the business. Engineering leaders should focus on solving hard business problems rather than nit-picking on technology choices. This sends a strong signal that business impact is more important than technical “superiority.”
To create a culture that encourages the team to focus on business impact, growth and promotions should be the result of solving problems that have a direct correlation to business impact. Be sure to recognize and reward engineers who spend hours fixing nasty bugs or resolving critical issues that have a huge business impact, especially if the work is not necessarily “sexy.” At company wide all-hands meetings, share graphs of performance or availability impact of different fixes. It is important to demonstrate that strong impact doesn’t just come from working on the “new shiny object.” Work involving performance upgrades, decreasing latency etc. may not be sexy, but they have massive customer impact.
Articulate the Path Forward by Creating an Engineering Ladder Early On
Create an engineering ladder/lattice that clearly articulates the path forward for engineers while driving the business’ success. The goal of a ladder is to set right expectations at every level. Without clear definitions of what constitutes success, decisions around promotions often boil down to years of experience. It is very important to debunk the commonly-held perception that tenure equals growth. Tenure should be relevant only insofar as it builds consistency and reliability in solving higher-order, more complex problems. The ladder should capture critical experiences and problem-solving abilities that individuals should acquire in order to progress through their careers. It’s a good framework that a manager can use to give an engineer things to do to get to the next level. It also helps when you hire people from outside. The interviewing process should be tailored to evaluate for candidates at different levels using the same ladder.
One thing to always keep in mind is to create both a “technical” and a “managerial” track. We wanted to eliminate perverse incentives for folks to become people managers. Senior individual contributors can make as big or even bigger of an impact than their managers and be rewarded accordingly. In fact, I would argue that becoming a manager is like changing your role. You are restarting your career from ground zero. Everything you learned as an individual contributor will help you in your managerial role, but may not be your primary responsibility as a manager. As you build the ladder, it’s important to keep things flexible. A person who becomes a manager may later decide that management is not for her. Encourage a lattice structure where a person on the managerial track can decide to move to the technical track, and vice versa.
Finally, a Note on Culture
I have used the word “culture” very lavishly throughout this post. Culture is not about putting up a poster of your core values on a wall. It’s about how you use those values to make decisions every day. Culture defines how people act day in and day out — it’s how individuals act when no one is watching. It reflects in how you celebrate success, reward employees and behave in conflict scenarios.
So, make sure you spend time on building your culture. Just like leadership, building the right culture takes time and patience. There are no tricks or shortcuts to culture-building. The best thing about culture is that you will know when you have built the right one. And it’s rewarding to see when the entire organization feels aligned, motivated, connected and cohesive!
Originally published at www.paxos.com on October 24, 2018.