I Turned a Free API Into a 00/Month Side Project

I Turned a Free API Into a $200/Month Side Project

The API existed. People needed it. They just didn't want to figure it out themselves.

The Origin Story

Last year I was building a crypto tracking dashboard. I needed real-time price data, signals, and historical performance metrics.

The problem? Every exchange has a different API format. Some use REST, some WebSocket, some GraphQL. Authentication varies wildly. Rate limits are all over the place.

I spent two weeks just normalizing data from 5 exchanges.

Then it hit me: if I need this, other developers do too.

What I Built

A simple API wrapper that:

1. Aggregates data from multiple sources into one format
2. Caches responses (so you don't hit rate limits)
3. Handles authentication automatically
4. Returns clean, consistent JSON

Before (what developers do):
┌─────────┐  ┌─────────┐  ┌─────────┐  ┌─────────┐
│ Binance │  │Coinbase │  │ Kraken  │  │ Bybit   │
│   API   │→ │   API   │→ │   API   │→ │   API   │
└─────────┘  └─────────┘  └─────────┘  └─────────┘
     ↑ Different auth    ↑ Different formats    ↑ Different limits

After (my service):
┌──────────────────────────────────┐
│         My Unified API           │
│                                  │
│  GET /api/v1/prices?symbol=BTC  │
│  → { "price": 67432, ... }      │
│                                  │
│  GET /api/v1/signals             │
│  → [{ "symbol": "ETH", ... }]   │
└──────────────────────────────────┘

The Tech Stack (Keep It Simple)

// server.js — Express.js, ~200 lines of code
const express = require('express');
const axios = require('axios');
const NodeCache = require('node-cache');

const app = express();
const cache = new NodeCache({ stdTTL: 300 }); // 5-min cache

// Single endpoint that does everything
app.get('/api/v1/signals', async (req, res) => {
  const cached = cache.get('signals');
  if (cached) return res.json(cached);

  const signals = await generateSignals();
  cache.set('signals', signals);
  res.json(signals);
});

app.listen(3001, '127.0.0.1');

Total codebase: ~500 lines including tests
Development time: 3 days (one weekend + one evening)
Hosting cost: $0 (runs on existing VPS)

How I Got My First Customer

I didn't market it. I didn't post on Product Hunt. I didn't run ads.

Here's what actually happened:

Step 1: Made It Useful for Myself First

I built it because I needed it. That meant:

• The API design was natural (it solved my own problem)
• The documentation was honest (I wrote what I wished existed)
• The edge cases were real (I'd already hit them)

Lesson: Don't build for an imaginary customer. Build for yourself, then see if others have the same problem.

Step 2: Mentioned It in Passing

I dropped a comment on a Reddit thread about crypto APIs:

"I built a unified signal API that handles all this. DM me if you want early access."

Result: 3 DMs. One became a paying customer.

Step 3: Set Up Pricing That Scales

I used a simple three-tier model:

Plan	Price	Requests/hour	Features
Free	$0	10	Basic signals, 1h delay
Basic	$9/mo	50	Real-time, email alerts
Pro	$29/mo	200	Everything + webhook + priority

Why this pricing works:

• Free tier removes friction (try before buying)
• Basic covers casual users (impulse purchase territory)
• Pro covers serious traders (high margin)

Key insight: The free tier isn't free for me. It costs almost nothing to serve (cached responses), but it brings in leads who upgrade later.

What Made Money (And What Didn't)

Revenue Timeline

Month 1: $0     (building, no customers)
Month 2: $29    (1 Pro customer from Reddit)
Month 3: $47    (1 Pro + 2 Basic)
Month 4: $38    (1 Pro cancelled, 1 new Basic)
Month 5: $66    (2 Pro + 2 Basic)
Month 6: $87    (steady growth from word of mouth)

6-month total: $267

Not life-changing. But for ~500 lines of code and zero marketing spend? Pretty good ROI.

What Didn't Work

1. Twitter promotion: 0 signups from 500 impressions. My followers are other devs, not traders.
2. Blog post: Got traffic, got GitHub stars, got zero paying customers.
3. Cold emails: Messaged 20 crypto newsletter writers. 2 replies, 0 conversions.
4. Free tier abuse: One guy wrote a script hammering the free endpoint. Had to add rate limiting.

What Did Work

1. Reddit comments: Casual mentions in relevant threads. 80% of customers came from this.
2. Word of mouth: One customer recommended it to a friend.
3. GitHub README link: People found the open-source core, wanted the hosted version.

Technical Lessons

Caching Is Everything

Without caching, I was hitting exchange APIs 1000+ times/day. With caching:

// Before: 1200 API calls/day → rate limited constantly
// After: 288 API calls/day (every 5 min) → never rate limited

const cache = new NodeCache({ stdTTL: 300, checkperiod: 60 });

async function getCachedOrFetch(key, fetchFn, ttl) {
  const cached = cache.get(key);
  if (cached) return cached;

  const value = await fetchFn();
  cache.set(key, value, ttl);
  return value;
}

Cost savings: My server handles 10x more requests with the same API quota.

Graceful Degradation

When one exchange API goes down (and they do), don't fail completely:

async function getPrice(symbol) {
  const sources = [binance, coinbase, kraken];

  for (const source of sources) {
    try {
      return await source.getPrice(symbol);
    } catch (err) {
      logWarning(`${source.name} failed: ${err.message}`);
    }
  }

  // All failed — return last known good price
  return cache.get(`price:${symbol}`) || null;
}

This single pattern prevented 99% of "your API is down" complaints.

Auth Done Right

// NEVER hardcode keys
const config = {
  binance: { key: process.env.BINANCE_KEY },
  coinbase: { secret: process.env.COINBASE_SECRET },
};

// Validate at startup
Object.entries(config).forEach(([name, cfg]) => {
  if (!cfg.key && !cfg.secret) {
    console.warn(`Missing credentials for ${name}`);
  }
});

If I Started Today, What I'd Do Differently

1. Start with a waitlist. Collect emails before building. Validate demand first.
2. Use Stripe from day one. I started with manual PayPal invoices. Lost 2 customers who didn't want to bother.
3. Write better docs. My biggest support burden is "how do I call your API?" A good OpenAPI spec would save hours.
4. Build the paid features first. I built everything free, then tried to add paywalls. Should've been reverse.
5. Add usage dashboards. Customers love seeing their own stats. Builds engagement and retention.

The Bigger Picture

An API wrapper isn't a groundbreaking product. It's not going to get you on TechCrunch.

But here's what it IS:

• Real revenue from day one (not "we'll monetize later")
• Skills you'll reuse (API design, caching, auth, billing)
• A portfolio piece that shows you can ship and maintain a product
• Understanding of what customers actually pay for

Every SaaS starts somewhere. Mine started with 500 lines of code and a Reddit comment.

---

Building something similar? I'd love to hear about it. Drop a comment or find me on Dev.to.

If you're interested in the unified signal API I mentioned, it's live at https://agentvote.cc/signal/ — free tier available.