11 months at a startup
Reflecting on my time at IB, some of the key lessons I learnt (in no particular order)
- Use Jira
Thats it! use jira if you have more than 1 team. Give the team a goal using epics, track all feature rollouts, bugs, contributions per employee, etc.
- Employee Morale and Churn
Startups go through waves. Sometime it feels like you are killing it. Close a new customer, launch a new product, land a great hire, etc. Sometimes it feels like outright dumpster fire. Inevitably, this leads to some employees feeling low and will quit. IB is no different, however one thing that helps a lot is if the core team, especially eng is stable. IB has all key managers stay through these cycles for many years and plan to stay for many more. This creates stability as foot soldiers churn is not as determinental as commander churn. An orthongal point is you need to identify and fire useless people and more importantly detractors fast.
I always felt the interview process is broken. Folks get asked standard questions which most of the times have no relation to work. I have developed some appreciation for it, where now I feel it is a resonably good thresholding factor. If you don’t do well in interviews on basic questions, you are likely not good at logic. That said, one thing missing in interview systems if compatibility factor - things like is the said person hard working or is fun to work with.
- Building an Institution
Small internet business that are quite profitable trade at 1-3x revenue. Many companies trade at 20x+. Investors find comfort even when investing in these companies as these are institutions. They are sort of like mini-governments, just faster and will chug along and continue to grow. Building a franchise like this requires defining processes and roles with clearly defined responsibilities. This makes it easy to scale. For eg, in the initial days, we had largely just eng with eng folks putting on many hats. Now there is PRE, Customer Success, Sales Eng, etc. Before we had a vaguely defined OKR with no open accountability and random items getting added an removed. Now we have a well defined OKR that is publicly advertised and status updates posted againt it. The former works for small teams but is chaotic and doesn’t scale. The later is slower but adds structure and is scalable.
- Accelerate the Future & Reinvent
This point is quite hardest to explain as it has so many facets. At a high level, this is what Elon refers to as speed to innovation or why AWS is probably the greatest software enterprise. Will expand on this point at a later date. Brief blurb: There is > 2x value in getting something that would be ready in 12 months out in 6 months. For example, if certain product is going to taking 12 months to ship with 2 engineers and process is scalable, is possible hire another 3 and try to ship it in 6 months.