You will find advice online that says to spend a certain percentage of revenue on marketing. It gets repeated because it is easy to say. It does not help you, because your margins, your sales cycle and your starting position are not average.
Start from the answer you want
Work backwards. Write down four numbers, using rough estimates if you have to.
- How many new customers you want each month.
- What share of enquiries becomes a customer. Your best guess is fine.
- How many enquiries that means.
- What you could afford to pay for one new customer, given what you earn from them.
An illustration, not a benchmark: if you want ten customers, and one in five enquiries becomes a customer, you need fifty enquiries. If you can afford to pay a thousand rupees to win a customer, the outer limit is ten thousand rupees of advertising for the month. Change the assumptions and the answer moves. That is the point: the exercise makes your assumptions visible.
Then spend less first
Almost nobody knows their real conversion rate before they start. So the first budget is a test. Size it to buy enough enquiries to learn something. That is usually more than people want to spend and less than they fear. Keep the test on one channel and one offer. Splitting a small budget across four platforms gives four results too small to read.
Not everything is ad spend
A marketing budget includes more than advertising. A page that converts, tracking that works, the reels and photographs you will need, and the time of the person who answers enquiries all cost something. In the early months, it is often better to put a share of the budget into the page and the follow-up than into more clicks.
Signs the budget is wrong
- Too low: the number of enquiries is too small to draw any conclusion, so every week is a coin flip.
- Too high, too early: you have not fixed the page or the reply speed, and each rupee is wasted at the same rate.
- Spread too thin: many channels, none big enough to learn from.
A fuller illustration for a service business
This is invented arithmetic, to show the method. A small service business earns a certain profit from each new customer and closes about one in five serious enquiries. To gain five new customers in a month, it needs about twenty five enquiries. If the business is happy to pay up to a certain amount to win each customer, that sets the outer limit for advertising in the month. If the first two weeks show that enquiries cost more than the limit allows, the options are to improve the page, the offer or the follow-up, or to accept that this channel does not suit the business. It is not to spend more.
Budget in phases
| Phase | Focus | What the money is for |
|---|---|---|
| First month | Foundations | A page that works, tracking, and a small test on one channel |
| Months two and three | Learning | Enough spend to see patterns, and changes based on them |
| After that | Scaling what works | More budget on what shows results, and less on what does not |
Warnings from vendors
- A firm that tells you a percentage of revenue you must spend without asking about your margins.
- A promise of a minimum number of leads for a given spend.
- A request to commit a large budget before any test has been run.
Review on a schedule
Decide when you will look at results and what would make you spend more, less or stop. Write it down before the campaign starts. Decisions made in the middle of a slow week are rarely good ones. Our piece on how much data you need before judging an ad covers the reading part.
If you would rather do the working-backward exercise with someone, that is what our diagnosis is for. Ad spend is paid to the platform and is separate from any fee.