Almost everyone approaches this from the wrong end. They decide what they can afford to lose, spend it, and then try to work out whether it went well.
The problem is that a number chosen that way has no relationship to whether the advertising can succeed. Spend too little and you will get a result too small to learn anything from, then conclude that ads do not work for your business.
There is a better sequence, and it starts with arithmetic rather than budget.
How do I work out what I can afford to spend?
Work backwards from what a customer is worth to you.
Three numbers. What is your average sale worth in profit, not revenue. How many of those does a typical customer buy over the time they stay with you. And what share of that profit are you willing to spend to acquire them.
Say a customer is worth forty thousand rupees in profit over two years, and you are willing to spend a quarter of that to win one. You can afford ten thousand to acquire a customer.
Now the only question that matters is whether advertising can deliver a customer for less than that. Everything else is detail.
Most small businesses have never calculated this, which is why ad budgets tend to be decided by feeling. Feeling is a poor guide, because the right number varies enormously by industry and yours is not intuitive.
What is the minimum worth spending?
Enough to produce roughly fifty conversions in a reasonable window, which for most small businesses means a few hundred dollars a month at minimum.
This is not a platform requirement, it is statistics. Below about fifty conversions the campaign cannot optimise properly and you cannot tell signal from noise. Ten leads in a month tells you almost nothing, however good or bad they look.
The practical consequence: it is better to run one campaign properly for two months than four campaigns weakly for two weeks each. Spreading a small budget across several tests is the most common way small advertisers learn nothing at all.
If you cannot afford the minimum for a real test, do not run a small one. Spend the money on something else and come back when you can.
Should I start small and scale up?
Start at the smallest amount that can produce a real answer, then scale on evidence.
Filling a bath with the plug out is a fair picture of underfunded advertising. You are spending continuously and never reaching a level where anything accumulates.
The sequence that works: fund one campaign properly for six to eight weeks. Measure cost per enquiry. Compare it to what you can afford. If it comes in under, increase spend gradually, roughly twenty percent at a time, and watch whether the cost holds. If it comes in over, change something substantial rather than spending more.
Sudden large increases usually push cost per enquiry up, because the platform has to find new audiences quickly. Gradual increases let it hold.
How should I split budget between Google and Meta?
They do different jobs, and the split depends on whether people are already looking for you.
Google search reaches people actively looking. Someone typing “web designer near me” has already decided they want one. High intent, higher cost per click, and usually a better cost per enquiry.
Meta reaches people who were not looking. Lower cost per click, much lower intent, but far better reach for something people did not know they needed.
If your service is one people search for, start with Google. If it is one people have to be shown, start with Meta. If both, start with the one where the buying intent is clearest and add the other once the first is working.
Running both from day one on a small budget usually means neither gets enough to prove anything.
What else should I budget for besides the ads?
The landing page, and the tracking, before any media spend.
Sending paid traffic to a page that was not built for it wastes a large share of what you spend. And without conversion tracking configured before the first rupee, you are buying clicks without knowing which ones became enquiries.
Both cost far less than a month of media. Both should exist first. This is the most common sequencing mistake in small business advertising, and it is expensive precisely because everything appears to be working.
When should I stop?
When cost per enquiry sits clearly above what a customer is worth, and has not moved after a genuine attempt to fix it.
That is the whole test, and it is why the arithmetic at the start matters. Without a number to compare against, results feel good or bad rather than being one or the other, and campaigns get stopped when they were working or continued when they were not.
Advertising is not a thing that works or does not. It is a purchase, at a price, and the only question is whether the price is one you can pay.
We set up tracking before spending anything, so the numbers above are ones you can actually see. Tell us what you are running.
