Embracing the 3-Month Rule: A Pragmatic Approach to Development in Startups
In the world of startups, entrepreneurs often hear the mantra from Paul Graham to “do things that don’t scale.” However, putting this advice into practice, particularly in the realm of coding, is rarely discussed. After eight months of building my AI podcast platform, I’ve developed a practical framework: any unscalable solution gets a three-month trial period. In that time, it must either demonstrate its worth and be developed further or be phased out altogether.
As engineers, we are naturally inclined to design scalable solutions from the outset. We think of complex architectures, microservices, and distributed systems—essentially, the infrastructure that allows for millions of users to engage. While these strategies are essential for larger corporations, they can lead to unnecessary complexity in a startup environment. My experience has shown that chasing scalability too early can be an expensive delay, particularly when optimizing for users that don’t exist yet. By adhering to my 3-month rule, I’ve learned to write straightforward, albeit “messy,” code that can be implemented quickly. This approach teaches me far more about what users genuinely require.
My Current Infrastructure Hacks: Simplifying for Success
1. Unified VM for All Operations
Currently, my entire platform operates on a single virtual machine, encompassing the database, web server, background jobs, and even Redis—all for just $40 a month. While this may seem imprudent due to the lack of redundancy and manual backups, it has led to invaluable insights about resource allocation in just a couple of months. Instead of over-engineering complex setups, I’ve learned that my “AI-heavy” platform actually peaks at 4GB of RAM, which means that had I deployed a Kubernetes solution, I would have been nurturing empty containers.
Every time the system crashes—an experience I’ve encountered twice so far—I gain tangible insights into what really fails. Unsurprisingly, these points are seldom what I initially anticipated.
2. Hardcoded Configurations for Agility
Instead of utilizing configuration files or environment variables, I’ve opted for hardcoded constants—like prices and maximum user limits—sprinkled throughout my code. Whenever I need to make adjustments, it involves a straightforward redeployment process. This might seem rudimentary, but the ability to quickly search my codebase for constants means that every change is thoroughly documented and reviewed, albeit by myself. Over the past three months, I’ve made only
