The list of speakers of Elevate 2020 was a seemingly endless list of CTOs and VPs of industry leading companies. My curiosity took over and I ended up attending the summit to learn from the best leaders in tech. I was satisfied by the quality of my notes and want to share them with you. Enjoy! Before we get started, I want to take some time to thank Plato HQ for organising this exquisite event.

Opening keynote

David Singleton, CTO of Stripe, opened the keynote with a talk about remote work. He showcases his remote engineering hub. It's a virtual office where async communication is king. The hub allows newcommers to introduce themselves, announcements to be made and incident response war rooms to be set up. He believes that shorter time between surveys is key in a remote world. It is also interesting that Stripe uses async video recordings as an alternative to text. These recordings are removed after a week to give that same casual sense of an in-person talk.

How do you prioritise user feedback? David replied that you should start with one single question: "Who are the users which will care about this the most?". It takes effort to find that very small group and even more effort to develop deep relations with these users. Only afterwards can you talk to them with what you think about doing and iterate with them together.

How do you resolve technical remote conflicts? David replied that each party should write down his opinion with reasoning. Often the conflict was a misunderstanding where data is missing. The techniques also scales. In case the problem does not resolve itself, you can trigger an unblocking process where the documents are send to a responsible individual which makes an executive decision. Do note that this only works if you use the disagree and commit principle.

Technical Leadership: How to be Intentional about Career Growth as an Engineer

Roojuta Lanani, principal engineer at Salesforce, kicked of the technical leadership track talking about growth. Growth can be put into three buckets: career, intellectual and personal.

The best way to tackle career growth is by talking with your manager. The conversation will be two-fold. Share your goals and ask for feedback. Managers want to help you grow, though they can not read minds. Be explicit about where you want to be in the next six months or year. She explains that asking feedback does not always need to go through official channels. Her advice is to make it a habbit to go for a coffee after having reached a minor milestone. That way your latest endeavours are still fresh in your minds and you can have deeper conversations about your strenghts and weaknesses.

Intellectual growth is all about the scope and impact of your projects. Be careful with waiting around and getting comfortable. Her observation is that the amount of available opportunities decrease as you grow. At that point, proactively search for the next, currently not yet known, big challenge and start solving it. Personal growth is about becoming a better human being. It is a great area to focus on when you feel stuck. One suggestion is to attend hackathons in which you play a different role to increase your empathy towards these roles.

Takeaways

Growth should be an active topic with your manager.
Growth has different, equally important areas. Check in from time to time that you are not neglecting some.

Product Leadership: The 7 Biases of Product Teams

Shreyas Doshi, product lead at Stripe, is well known for his Twitter presence. This talk was all about biases. Understanding biases is understanding that our minds are imperfect. The solution is to cultivate awareness by asking questions and growing a shared vocabulary around common flaws. Below are the names of each biase combined with a reflective question and a way to counter it. I highly recommend his latest Twitter thread which goes deeper into the topic.

Focusing illusion for products.

Am I overestimating the severity of the problem I am solving?
If yes, counter with deep customer empathy.

IKEA effect for products.

Am I attaching to much value to something I build myself?
If yes, counter with rigorous thinking (i.e. reasoning from first principles)

Bias-for-building fallacy.

Am I jumping to building software over solving a problem?
If yes, counter with velocity over speed.

Execution orientation fallacy.

Am I blaming circumstances?
If yes, counter with ambition and high agency.

Maslow's Hammer.

Am I choosing this tool for convenience/the wrong reason?
If yes, counter with context comes first.

Russian Roulette for products.

Am I neglecting risk management?
If yes, counter by a culture of prevention over heroics.

Authority Approval Bias.

Am I letting people in a position of authority instead of customers drive my proposals?
If yes, counter by constructive dissonance. Do not try to proof yourself right.

In the Q&A he remarked that you should aim for velocity in the right direction. It's not about writing code quickly. It's about understanding your customer quickly.

Technical Leadership: How engineers can influence other engineers & PMs

In this panel, three high-positioned employees from Google and AWS talk about influence.

Why does influence matter? Think about influence as leadership. You may not think that you are a leader, but you are. Within your own space you are your own leader. As time passes, you can help your peers and people who joined after you. This kind of leadership will be noticed by your management. You will take a driver seat in your own career and from there onwards be able to help others achieve that as well.

How do you build influence?

It is not about working more hours, you should not try to do everything yourself.
It is not about authority attached to a title.
It is building trust. You need to be humble and trustful.
It is having a seat at the table. Be an exquisite listener and understand your role at that table. Influence is output, you have to bring value to the conversation. Over time this output leads to trust.
It is learning the skill to get people excited to invest their time.
It is partnering with more influential people. Trust is transitional.
It is asking the right questions, not about technologies, but about the current problems.
It is understanding your brand. Cultivate it, bring it forward and let people leverage you.

Technical Leadership: Quantifying Quality

Keith Adams, Former Chief Architect at Slack, definitely had the most technical talk. He joined Slack when growth was great but users complained about the service being slow, buggy and unavailable. They even worried Slack was insecure.

The problem is that features are fun to play with and great maintenance is unnoticed. Feedback comes too late and management has to be convinced about the importance of maintenance to set aside budget.

The solution is the SPQR framework in which Keith tries to quantify quality. SPQR stands for security, performance, quality and reliability. It's all about the odds of having a good experience, which equals the chance of not hitting a problem. His formula shows that quality issues do not add up; they multiply up. It shows how over time a system can die by a thousand cuts.

Reliability and performance can be quantified by observability. Reliability is the uptime multiplied by the successful number of actions divided by the total number of actions. Performance are the fast enough number of requests divided by the total number of requests. It turns out that the exact threshold for fast enough does not matter much. When talking about to slow, we are not talking about 10ms to 50ms, but more along the lines of 3s and 5s. You should be careful about setting thresholds to aggresive as you will lose to much time.

Quality and security can be quantified by a bug database. While it might a bit odd that security is quantified this way, real incidents are too sparse and too late to be useful. Instead the real useful metric are research topics and known security issue databases. That way you can instead treat them as a bug. Once added to the database, a severity label indicates the probability to give an unacceptable experience. He also notes that while customer tickets are valuable, they undersample the real amount of bad experiences.

To conclude, multiplication is cruel. Reducing the probability of each problem increases the likeliness of hitting a good experience. This is all that matters in the end.

Closing keynote

The keynote was a Q&A dialogue between Sonali Sambhus and Jason Warner, CTO of GitHub.

What is the fundamental purpose of leadership? The fundamental purpose of leadership is to help people achieve something. A shared mission and objective. If you are really good at your job, you make something happen that would otherwise not have happened. It is really as simple as that. One hundred years ago and one hundred years from now this will stay the same. Technology might change, but core lessons still apply. How do you actively manifest something in the world? For me it sounds very simple, communicate it. Why you do it, for who you do it. Lay down the objectives and inspire people to achieve them. You need to find people who have the same passion of you. You will have many people who you will not be able to inspire as you just have different passions.

What are leadership skills valued five years from now? It's only recently that empathy has raised to the top. We realised that it are people who create value. The skills most values are the combination of empathy and getting things done as a leader. Building trust is crucial in keeping the balance between empathy and authority. Culture is about having a shared mission and inspiring your objectives. You will fail if you reside at one side of the spectrum. For example, you can be the most achieving person but burn people out quickly, and you will fail. You can also be the most caring person and never achieve something, and you will fail. You need to find the balance.

What would you tell to your ten-year past self? Trust your instincts more. Being right in retrospect is the worst feeling ever. On top of that, don't let these well established leaders scare you off, everyone wakes up from time to time with uncertainty and a feeling that they are messing up.

What would you advise to move up the career ladder? Sonali explains that a ladder is not the right analogy, think of it as a career lattice. There is a lot of room to move around and to make jumps. It is all about operating at a higher level. Be brave, make big bold leaps. Jason continues that you have to look to who you are reporting. Your manager will have a big impact on your promotion. Also learn how to motivate your team and from time to time direct them. At the end of the day you will not be able to cheat these things. You need time to learn them. Finally, he warns that the difference between manager, director and VP are dramatic. Vice Presidents operate with huge budgets and you won't know if you are right for years. Be introspective what you are good at and what you are lacking. Improve your flaws but double down on your strenghts. They are your super powers.

What is the biggest threats to technology in the current pandemic? Regulation will increase and compliance will become harder than ever to navigate. Technology will also keep becoming more complex. For example, Kubernetes is a fantastic software but it is complex and tricky to navigate.

In his closing words, Jason reiterated the importance of deep introspection on yourself. Fast forward five year from now. What are your superpowers? How can you make them stand out? What are your weaknesses? How can you reduce their impact? Never lose that growth mindset.