"How much should I spend on ads?" is one of the questions business owners ask most, and one of the most often answered wrongly. The common answers ("start with $50 a day" or "10% of sales") know nothing about your business. The right budget is calculated from your own numbers.

Step one: calculate the maximum customer cost you can afford

If your average order is EGP 600 and your margin is 35%, the most you can pay for one order without losing money is EGP 210. If you sell cash on delivery, adjust for delivery rate. Full details in break-even ROAS.

That number is the foundation: every budget decision is compared against it.

Step two: a campaign needs data to learn

The algorithms at Meta, TikTok and Google need a reasonable number of conversions to know whom to look for. A campaign getting one or two conversions a week stays in guessing mode.

So ask: how many conversions a week does the campaign need to learn? Then:

Worked example

Say the expected cost per order in your market is around EGP 150, and you want to give the campaign a reasonable number of orders each week, such as 20. The weekly budget is roughly EGP 3,000, or about EGP 430 a day.

If that is more than you can spend, you have two options:

  • Optimise for a nearer, more frequent event at the start (such as initiate checkout instead of purchase), watching the quality of results.
  • Reduce the number of campaigns and ad sets so the data concentrates in one place.

The biggest mistake is splitting a small budget across many campaigns, so none of them learns.

Step three: a testing budget, then a scaling budget

Split the start into two phases:

  1. Testing: a budget fixed in advance for a fixed period, whose goal is learning, not profit. Which message works? Which audience? What is the real cost per result? Decide in advance how much you are willing to spend to find out.
  2. Scaling: once you find what delivers a cost below your maximum, increase budget gradually on what works only.

When to increase the budget

  • When the real cost per result (delivered order, qualified lead) is clearly below your maximum.
  • When performance has been stable for a while, not for one day.
  • Increase gradually, because big sudden jumps can unsettle the campaign and send it back into learning.

When to stop or change

  • When the real cost per result stays above your maximum after enough data.
  • When the campaign brings cheap, poor results (orders that are not delivered, leads who do not answer). See reducing COD returns.
  • When the problem is the offer or the page, which more budget will not fix. See the landing page.

What about the management fee?

The ad budget is one thing, the fee for whoever runs the campaigns is another. If the ad budget is very small, a management fee may not make sense relative to it. In that case, start with an account review or consultation, and run the campaigns yourself until the budget grows.

The bottom line

A budget is not set by guesswork or a fixed percentage, but from two numbers: the maximum customer cost you can afford, and the number of results a campaign needs to learn. Start with a fixed testing budget and scale only what proves profitable. If you want help calculating your budget and a starting plan, see the freelance media buyer page.