The 3-Month Rule: A Pragmatic Approach to Building Unscalable Solutions in Startups
In the world of startups, the mantra “do things that don’t scale,” famously advised by Paul Graham, is often heralded. Yet, the implementation of this concept, particularly in coding, is not frequently discussed. After dedicating eight months to developing my AI podcast platform, I’ve devised a straightforward framework: every unscalable hack is granted a lifespan of three months. At the end of this period, the hack must either demonstrate its worth and be properly implemented or be discarded.
The startup Mindset: Embracing Imperfection
As engineers, we are typically conditioned to strive for scalable solutions from the outset. High-level design patterns, microservices architectures, and distributed systems are common in the lexicon of larger enterprises. However, for startups, this approach can lead to costly delays and unproductive optimizations. My three-month rule encourages a focus on writing straightforward, albeit “imperfect” code that is deployable and, more importantly, provides insight into user requirements.
Harnessing Infrastructure Hacks: Smart Choices
Here are a few of the unscalable strategies I have adopted and the rationale behind their success:
1. Single VM for Everything
I run my entire operation—database, web server, background jobs, and Redis—on one $40/month virtual machine, with manual backups managed locally. This approach may seem reckless at first glance, but it has allowed me to gain valuable insights into my actual resource needs in just two months. I’ve discovered that my platform peaks at 4GB of RAM, which means the complex Kubernetes scheme I nearly deployed would have involved unnecessary overhead.
2. Hardcoded Configuration
Rather than creating configuration files or utilizing environment variables, I have hardcoded constants throughout my codebase:
python
PRICE_TIER_1 = 9.99
PRICE_TIER_2 = 19.99
MAX_USERS = 100
AI_MODEL = "gpt-4"
This method makes it incredibly easy to search for configuration values and allows for quick redeployment. I’ve adjusted these values only three times in three months, resulting in a mere 15 minutes of redeployment time instead of a week’s worth of engineering effort.
3. SQLite in Production
Believe it or not, I am using SQLite for a multi-user web application. With a database size of just 47MB, it efficiently
