Embracing the 3-Month Rule: A Strategic Approach to Unscalable Solutions in Tech
In the tech startup world, the conventional wisdom advocates for “doing things that don’t scale,” as famously advised by Paul Graham. However, while this philosophy is widely acknowledged, the practical execution of it—especially in the realm of coding—remains a less-explored territory.
Over the past eight months of developing my AI podcast platform, I’ve crafted a straightforward yet effective framework: every unscalable approach gets a lifetime of just three months. By the end of this period, each solution must either demonstrate its value well enough to warrant further investment or be let go.
As engineers, our training often nudges us toward scalable solutions from the outset—think design patterns, microservices, and distributed systems. Such architectures are impressive and can handle millions of users, but they represent the mindset of larger enterprises.
In contrast, focusing too much on scalability in a startup can tether you to burdensome procrastination, optimizing for theoretical users and addressing potential problems that may never arise. My 3-month rule encourages me to adopt simple, direct, and sometimes ‘messy’ coding practices, enabling me to identify the genuine needs of my users.
Current Infrastructure Hacks: Why They Make Sense
1. Unified VM for Everything
For just $40 a month, I’ve consolidated my database, web server, background jobs, and Redis on a single virtual machine. There’s no redundancy and backups are manual.
Why is this a smart move? In two months, I gleaned more about my resource requirements than any elaborate planning document could provide. I discovered that my “AI-heavy” platform only peaks at 4GB of RAM. The complex Kubernetes setup I initially considered would have just been a management overhead for empty containers.
When failures occur (and they have, twice), they reveal invaluable insights into actual problems—none of which I anticipated.
2. Hardcoding Configurations
python
PRICE_TIER_1 = 9.99
PRICE_TIER_2 = 19.99
MAX_USERS = 100
AI_MODEL = "gpt-4"
All configurations are hardcoded within the codebase itself, with no separate config files or environment variables. Each change necessitates a redeployment.
The beauty is in its simplicity; I can quickly search for any config value across my codebase. This approach allows me to review changes through
