Kevin Systrom (Instagram) and Lex Fridman on Product

Videos · Startups · Source

Scalability and PMF

Don't obsess about scalability at first. It's very rare that startups can't scale, you can just be inefficient and throw money (more instances) at the problem if needed

But obsess about product-market fit. Obsess about growth, obsess about user retention cohorts and user love

Product experience

Hid the latency of compute-heavy tasks by a) processing them in the background, and b) pre-loading the stuff that comes after the task has been processed

For example, even though applying a filter took around 3 seconds, users thought it was fast because a lot of the other stuff around uploading it etc happened in parallel

Focus obsessively about what job your product does for your user, not just what features are hot rn

Optimizing for speed of iteration vs creating the "perfect" product

Being obsessive about testing early on was helpful. Tests helped them move faster, specially when they were releasing new features

Until the first 10 million, it was a simple Django app with Postgres and Redis. Didn't try to overcomplicate things. No Machine Learning

Monetization

You don't have to know exactly how you'll make money. But have an inkling

The best places to play are where your competitor cannot pivot because they're structurally not setup to be there

Changes in recommendation systems

Recommendation systems from social signals and for social points vs recommendations systems about the content and about user engagement

Search and discovery vs follows