30 Days at the Triangle Factory
The following is an attempt to answer two questions I've been asked lately:
What are you actually doing at Vercel?
How’s it going?
Vercel
Describing Vercel to other people feels like trying to explain an alien planet.
It’s hands down the fastest-moving, fastest-shipping, highest trust/autonomy company I’ve ever seen with an incredible density of talent. And at Vercel — this works, but I don't think it would at most other places.
Internally, we work and communicate in public as much as possible, and Malte’s motto “you can just ship things” is a way of life. I haven’t gone through an actual planning process, but there aren’t OKRs so much as directional guidance on where we need to go and a high level of trust that the teams will get there (as well as check-ins along the way).
For someone who is responsible for shepherding humans, this could be a miserable environment. But it isn’t. It’s amazing.
Trying to enable permissionless innovation and development at scale, without slowing down, is an incredible problem to own.
My Charter
Officially, I’m “Chief of Staff, Product & Engineering" at Vercel.
Chief of Staff (CoS) roles are highly varied in the industry, and even at Vercel the three of us have materially different mandates.
As of now, I have two focus areas:
Improve velocity and alignment for the Engineering, Product, and Design (EPD) Org.
Unburden and provide leverage to CPO Tom Occhino and CTO Malte Ubl.
Given how broad those areas are, I recognize this may not be helpful clarification. But that’s kind of the point. My job is to do whatever it takes to make Vercel better, to make EPD more efficient and effective, and to improve how we work and work together. I'm given problems, not projects, and I’m trusted to do whatever I can to fix them.
In “Being Hired To Care” I talk about the fun challenge of this way of working:
The empowering aspect of this is that without a lane or a specific role that you’re supposed to play, you can draw from your full toolkit. This quickly shifts you from playing sheet music to jazz… When you’re hired to care you have to trust yourself a bit more and ruthlessly prioritize what you believe will move the business forward, since people will rarely specify how you should contribute.
I think the fact that I’m hired to do this is a big reason why I got excited about joining. It didn’t feel like a massive departure from the work I was doing as Yet Another Studio, instead it was a chance to practice it deeper within a single organization.
On my first day, I shared the above image with the CPO (lovingly) and, so far, I feel empowered to do whatever I can to improve things.
ShipOps
Tactically, my work is very much split along the two focus areas I mentioned above. In the former, I’m responsible for what I’ve coined “ShipOps” which is the set of rituals, artifacts, and systems we have to enable everyone to just ship things. This is a mix of things that may live under product-, research-, or biz-ops umbrellas in other companies (along with things that have no home):
Rituals – the common meetings and rhythms of our work (i.e. weekly leadership meetings, planning cycles, etc)
Artifacts – the common objects that EPD interacts with (i.e. certain Notion docs, Slack channels, etc)
Systems – the social and technical systems EPD needs to function (i.e. integrations across tools, automations, etc).
We’re essentially building the F1 team so that every engineer feels like Lewis Hamilton and can focus on driving the car. Some of this is communication work, some of it is coordination work, and some of it is building.
As I mentioned above, Vercel is a fast-shipping, high trust company, so ShipOps is and will be critical to ensure the right things are going out to the right people in the right ways. One of Vercel’s core values is ITG or “iterate to great(ness)” and a lot of what we do in this space takes this approach of improving a little bit each week.
Some of the specific challenges I’m tackling right now:
In a world where “you can just ship things” what is the minimum necessary process to keep track of what’s going out and ensure that partner teams are properly informed and involved?
What does good change management look like in a dynamic and reasonably-autonomous organization?
Which of our existing systems will help us as we grow/scale and which of them will hinder us?
Leverage
My other focus is providing leverage for leadership, which looks different every day. It involves being their representative in meetings (planning, executing, follow ups, etc.), fielding requests, getting ahead of their schedule, and a whole host of other dark arts I’m picking up. This is the highest variance part of the job, since the ShipOps world is much more straightforward to plan for. But that makes this the highest contributor to my growth (so far).
Given Vercel’s growth trajectory, a lot of this work is me looking at how the CPO and CTO are spending their time and asking “If we hired you tomorrow, would you own this?” and rebalancing accordingly.
How It’s Going
From my perspective? Great. (My laptop still works, so probably also not terrible from Vercel’s side.)
I had a lot of questions going in about how this would feel. I wasn’t sure I was ready to “give up” the studio. I hadn’t worked as a full time employee at a company in almost five years. I have never written any React or Next.js.
As I mentioned in my Year 04.5 review:
If it was not already clear, I’m so grateful to be able to work with people I respect on things I care about and earn a living. Almost nothing on my journey so far has been consistent except for that fact.
This throughline has continued at Vercel. The people who I already knew there are amazing to work with and the people that I’ve met have been just as good. And like I said above, this doesn’t feel that different from the work that I was doing. But now I can do it in a more focused capacity and go deeper.
Last year I had the chance to do deep projects at both Mozilla and OpenAI, and, in both cases, I think I would have done a better job if I had been an employee because of the access and context I would have had. That is overwhelmingly true at Vercel too. This kind of “professional fixer” work would not be possible from the outside (and it’s a lot more fun on the inside).
One of the best things I’ve done since joining is start a #behzod-notes channel where I post observations each day. Vercel has a culture of truth seeking, and there’s no faster way to being right than being wrong in public.
I use this channel as a mix of journal and newspaper, home for my own reflections and organizational anthropology. The channel has over 100 people now, which has resulted in dozens of valuable conversations that help me learn faster (and hopefully will help Vercel do better) as well as me being tapped for a few initiatives. I continue to believe that writing in public has an insanely high ROI and this has definitely been the case at Vercel.
(Un)Learning
Perhaps more important than what I am learning at Vercel is what I am un-learning. I’m having a decade of corporate culture, practices, and beliefs challenged and nudged out of me as I figure out how to do things The Vercel Way. On Day 12, I noted that “Vercel felt uncomfortably fast at first, now it just feels fast.” I’m noticing this kind of change in many aspects of my work.
When I was hired, one of the pieces of feedback I got was essentially “you know that we need better systems, but you want to build bespoke ones rather than force-fit existing ones.” I’m trying to embrace that perspective in everything I do here, asking “where did this come from?” “who built this?” and “what is helpful/hurtful about the way this is right now?”
Curiosity, far more than judgement, will be the key to any kind of success I’ll have at Vercel.
Thank you to Tom Critchlow, Jasdev Singh, Parteek Saran, and Michael Nguyen for their feedback.