Planning a Campaign Budget

"How much should I spend on ads?" is the question every small business owner asks, and "it depends" is the honest but useless answer. A better answer is a calculation: start from the number of customers you need, work backwards through your funnel, and see what that costs. Then compare the cost with what a customer is worth.

In this lesson you will build a budget backwards from a goal, add a testing reserve and seasonality, split it across channels, and make a calculator the whole team can use.

Work backwards from the goal

Suppose a computer training institute in Rohtak wants 20 new students next month. Its own records give the funnel rates; the cost per click is an estimate from a short test campaign or from Google Ads Keyword Planner's bid ranges.

import math

goal_students = 20
lead_to_student = 0.25      # 1 in 4 enquiries joins (from last 3 months of records)
click_to_lead = 0.06        # 6 in 100 landing-page visitors enquire (from GA4)
cost_per_click = 22         # rupees, estimate from a test campaign

leads_needed = math.ceil(goal_students / lead_to_student)
clicks_needed = math.ceil(leads_needed / click_to_lead)
budget = clicks_needed * cost_per_click

print(f"Leads needed:  {leads_needed}")
print(f"Clicks needed: {clicks_needed}")
print(f"Ad budget:     ₹{budget:,}")
print(f"Cost per student: ₹{budget / goal_students:,.0f}")

Leads needed:  80
Clicks needed: 1334
Ad budget:     ₹29,348
Cost per student: ₹1,467

math.ceil rounds up, because you cannot get 0.3 of a student. Now the owner can make a real decision: if a student is worth ₹12,000 in fees, spending about ₹1,500 to get one may be fine. If the course fee is ₹2,000, it is not, and the plan must change (better conversion, cheaper channels, or a smaller goal).

Where do the numbers come from?

Every input should come from your own data whenever possible:

• Lead-to-customer rate: count enquiries and new customers in your register or CRM for the last few months.
• Click-to-lead rate: GA4 key events divided by sessions from the paid channel.
• Cost per click: your past campaigns; for a new campaign, Keyword Planner's estimated bid ranges or a small test budget for two weeks.

If you have no data at all, run a small test budget first, for example two to four weeks on one channel, and only then plan the full month. Planning from someone else's "average CPC" is guessing.

Test, scale and reserve

A sensible budget is split into parts. A common rule of thumb (not a law) is:

Keep a small reserve too. Leads sometimes arrive faster than expected, and a campaign that works should not stop on the 20th because the money ran out.

Seasonality

Demand changes through the year: admissions season for institutes, wedding season for photographers and caterers, Diwali for sweets and gifts, summer for AC repair. Google Trends and your own sales history show the pattern. This program spreads a yearly budget by monthly weights:

yearly_budget = 3_00_000      # rupees, written Indian-style with underscores
weights = {                   # relative demand from last year's enquiries
    "Apr": 3, "May": 3, "Jun": 4, "Jul": 3, "Aug": 2, "Sep": 2,
    "Oct": 1, "Nov": 1, "Dec": 1, "Jan": 2, "Feb": 2, "Mar": 2,
}
total = sum(weights.values())
for month, w in weights.items():
    print(f"{month}: ₹{yearly_budget * w / total:>9,.0f}")

Apr: ₹   34,615
May: ₹   34,615
Jun: ₹   46,154
Jul: ₹   34,615
Aug: ₹   23,077
Sep: ₹   23,077
Oct: ₹   11,538
Nov: ₹   11,538
Dec: ₹   11,538
Jan: ₹   23,077
Feb: ₹   23,077
Mar: ₹   23,077

Python accepts 3_00_000 because underscores in numbers are ignored, so you can group digits the Indian way (lakh and crore) for readability; the program still prints them in the international style.

A calculator for the team

Not everyone will run Python. This page does the same backward calculation in the browser. Change the inputs and the result updates.

<!DOCTYPE html>
<html lang="en">
<head>
  <meta charset="utf-8">
  <title>Campaign budget calculator</title>
  <style>
    body { font-family: system-ui, sans-serif; margin: 1.5rem; max-width: 28rem; }
    label { display: flex; justify-content: space-between; margin: .4rem 0; }
    input { width: 7rem; }
    #result { margin-top: 1rem; background: #fff7ed; padding: .8rem; line-height: 1.6; }
  </style>
</head>
<body>
  <h1>Budget calculator</h1>
  <form id="f">
    <label>Customers wanted <input name="goal" type="number" value="20" min="1"></label>
    <label>Lead to customer (%) <input name="close" type="number" value="25" min="1"></label>
    <label>Click to lead (%) <input name="cvr" type="number" value="6" min="0.1" step="0.1"></label>
    <label>Cost per click (₹) <input name="cpc" type="number" value="22" min="1"></label>
    <label>Value of a customer (₹) <input name="value" type="number" value="12000" min="1"></label>
  </form>
  <div id="result"></div>
  <script>
    const form = document.getElementById("f");
    const inr = (n) => "₹" + Math.round(n).toLocaleString("en-IN");
    function calc() {
      const v = Object.fromEntries([...new FormData(form)].map(([k, x]) => [k, Number(x)]));
      const leads = Math.ceil(v.goal / (v.close / 100));
      const clicks = Math.ceil(leads / (v.cvr / 100));
      const budget = clicks * v.cpc;
      const perCustomer = budget / v.goal;
      document.getElementById("result").textContent =
        `Leads needed: ${leads}. Clicks needed: ${clicks}. Budget: ${inr(budget)}. ` +
        `Cost per customer: ${inr(perCustomer)} (${(perCustomer / v.value * 100).toFixed(1)}% of a customer's value).`;
    }
    form.addEventListener("input", calc);
    calc();
  </script>
</body>
</html>

toLocaleString("en-IN") formats rupees in the Indian style, so 125000 is shown as ₹1,25,000 with the lakh comma. Try lowering "Click to lead" from 6% to 3%: the budget doubles. That one number shows why improving the landing page often saves more money than any bidding trick.

Costs beyond media

Ad spend is not the whole budget. Also plan for:

• Creative: photos, short videos, design.
• Tools: website hosting, email platform, scheduling tools.
• People: your time or a freelancer's, to manage campaigns and reply to leads quickly.
• Taxes and fees: check whether tax is added to ad spend in each platform's billing settings, and confirm the treatment with your CA.

Common mistakes

• Budget first, goal later. "₹10,000 on Facebook" without a target cannot be judged.
• Using someone else's benchmarks for CPC and conversion rates.
• Spreading a small budget across five channels. Each gets too little data to learn anything.
• No testing reserve, so nothing new is ever tried.
• Ignoring follow-up capacity. Doubling leads is pointless if nobody can call them back the same day.

Key points

• Plan backwards: customers wanted → leads needed → clicks needed → budget.
• Use your own conversion rates and test-campaign costs, not generic averages.
• Compare cost per customer with what a customer is worth before approving a budget.
• Split budget into proven, improve and experiment parts, keep a reserve, and follow seasonality.
• Include creative, tools, people and taxes in the full marketing budget.

import math

goal, close, cvr, cpc, value = 6, 0.10, 0.04, 30, 80_000
leads = math.ceil(goal / close)
clicks = math.ceil(leads / cvr)
budget = clicks * cpc
print(leads, clicks, f"₹{budget:,}", f"₹{budget / goal:,.0f} per booking",
      f"{budget / goal / value:.1%} of a booking's value")

60 1500 ₹45,000 ₹7,500 per booking 9.4% of a booking's value