Cloud Cost Basics

Build the mental model

A cloud bill can look intimidating at first glance, but the categories that typically get charged form a fairly small, consistent set.

• Compute — time a server runs
• Storage — volume of data kept
• Database — a managed database instance
• Bandwidth/Egress — data leaving to the outside world
• Requests — number of API calls
• Build minutes — time spent compiling and deploying
• Logs — log storage volume

The single most important distinction is fixed cost versus usage-based cost. A fixed cost charges roughly the same amount regardless of traffic. A usage-based cost scales with actual consumption.

COST BREAKDOWN MAP
------------------
COST BREAKDOWN MAP
---------------------

  Compute   ---\
  Storage   ----\
  Database  -----\
  Bandwidth ------>---[ TOTAL CLOUD BILL ]
  Requests  -----/
  Build     ----/
  Logs      ---/

  Each category can be FIXED (same cost every period)
  or USAGE-BASED (scales with consumption). Most bills
  mix both types across different categories.

Connect it to a real scenario

This example never touches a real pricing number — it's here purely to show the STRUCTURE of a cost bill.

It sets an illustrative placeholder unit cost for four categories — compute, storage, bandwidth, and requests — and computes a rough relative percentage breakdown.

• Cloud cost categories include compute, storage, database, bandwidth, requests, build minutes, and logs
• Understanding fixed cost versus usage-based cost matters more than any single number
• Free tier has a real usage limit and can be exceeded in production

Try the working example

function estimateCostBreakdown(usage) {
  // ILLUSTRATIVE placeholder unit costs only - NOT real provider pricing
  const illustrativeUnitCost = {
    computeHours: 0.05,
    storageGB: 0.02,
    bandwidthGB: 0.09,
    requestCount: 0.0000004
  };

  const rawCost = {
    compute: usage.computeHours * illustrativeUnitCost.computeHours,
    storage: usage.storageGB * illustrativeUnitCost.storageGB,
    bandwidth: usage.bandwidthGB * illustrativeUnitCost.bandwidthGB,
    requests: usage.requestCount * illustrativeUnitCost.requestCount
  };

  const total = Object.values(rawCost).reduce((sum, n) => sum + n, 0);
  const breakdownPercent = {};
  for (const category in rawCost) {
    breakdownPercent[category] = Math.round((rawCost[category] / total) * 1000) / 10;
  }

  return {
    estimatedTotalUnits: Math.round(total * 100) / 100,
    breakdownPercent
  };
}

const smallApiWorkload = {
  computeHours: 720,
  storageGB: 50,
  bandwidthGB: 200,
  requestCount: 2000000
};

console.log(JSON.stringify(estimateCostBreakdown(smallApiWorkload), null, 2));

{
  "estimatedTotalUnits": 55.8,
  "breakdownPercent": {
    "compute": 64.5,
    "storage": 1.8,
    "bandwidth": 32.3,
    "requests": 1.4
  }
}

(units are placeholders, not real pricing — this shows that compute and bandwidth drive most of the bill in this scenario)

Wikipedia: Cloud computing (pricing models section) — Cloud Providers & Platforms