How much should a startup spend on ads? A simple way to set your first budget

How to set a startup ad budget in India, the US, the UK or anywhere: work back from your goal, test small, then scale what pays back. No guesswork.

Short answer: don't start from a number someone else spends. Start from what one customer is worth to you, work out what you can afford to pay to get one, and set a small test budget that buys enough clicks to learn from. Scale only the ads that pay back. The same method works whether you spend in rupees, dollars, pounds or Singapore dollars.

Step 1: Know what a customer is worth

Before you spend anything, write down:

  • Average order value or contract value: what one customer pays you the first time.
  • Gross margin: what's left after the cost of delivering it.
  • Repeat value: how often they come back, if they do.

Together these tell you the most you could pay for one customer and still make money. That number is your ceiling, not your target.

Step 2: Set a target cost per customer

Pick a target cost per acquisition (CPA) well below the ceiling. Early on, many startups accept a higher cost per customer to learn, then work it down. Be honest about how long you can afford that.

Step 3: Work back to a test budget

Ads don't turn visitors into customers one for one. So work backwards:

  1. Visitors needed: how many people must visit to get one customer? If you don't know yet, your first test will tell you.
  2. Cost per click: look it up in Google Keyword Planner for search, or run a small Meta test.
  3. Test budget: enough clicks for a handful of conversions per campaign, so the results mean something.

A test that's too small just gives you noise. A test that's too big wastes money before you know what works.

Step 4: Split the test sensibly

  • One or two channels first. See Google Ads vs Meta Ads for startups to choose.
  • Most of the money on the main bet, a smaller part on one experiment.
  • Several ads per campaign. The ad itself decides most results, especially on Meta and YouTube. AI video ads make testing many versions affordable.

Step 5: Scale only what pays back

After the test, keep the campaigns whose cost per customer is under your target, and raise their budget in steps. Cut the rest. This loop is the core of performance marketing.

Does it change by country?

The method doesn't. The inputs do:

  • Cost per click differs a lot between markets. The same keyword can cost far more in the US, UK, Australia or Singapore than in India.
  • Conversion rates differ with price levels, trust and payment habits.
  • Competition is higher where more advertisers chase the same customers.

So if you sell in several countries, run a small test in each and compare cost per customer, not cost per click.

Mistakes that burn budget

  • Spending before tracking is set up.
  • Judging on clicks or likes instead of leads and sales.
  • Spreading a small budget across five channels.
  • Stopping a test after a few days.
  • Raising budgets too fast on a campaign that only just started working.

How we can help

We set up the tracking, plan the test, make the ads and run the campaigns for startups in India and worldwide, then report the cost per customer at every milestone.

All articles

Let’s chat and shape your brand.

Tell us where your startup is and where you want it to go.