Start with the floor rather than the target, because most Indian small businesses are sitting under it and do not know.
Your monthly floor on any paid channel is roughly thirty times your cost per lead.
Google’s own guidance is that automated bidding needs about 30 conversions in a 30 day window before it works properly, and 50 for target ROAS. Below that the system never leaves the learning phase, performance stays volatile, and you are paying for machine learning with nothing to learn from.
So if your cost per lead is Rs 500, your floor is about Rs 15,000 a month on that one platform. If it is Rs 1,200, your floor is Rs 36,000.
The Rs 8,000 a month “let us test it first” budget that half the SMEs I talk to are running is not a small version of a working campaign. It is a different thing that does not work, and no amount of skill fixes it.
That is the number that actually matters. The percentage benchmarks everybody quotes come second, and here is why.
What the benchmarks say, and what they are for
Gartner’s 2026 CMO Spend Survey puts the average marketing budget at 7.8% of company revenue. The CMO Survey run by Deloitte, Duke and the AMA puts it nearer 9.0%. Smaller companies run higher, with the smallest revenue band in those surveys reported at around 15.6%.
Those figures are worth knowing and they are a sanity check, not a method.
A percentage of revenue tells you what companies like yours happen to spend. It tells you nothing about what a customer is worth to you, what one costs to acquire, or whether the channel you are about to buy can even reach its floor. Two firms with identical revenue in the same category can have correctly different budgets by a factor of three.
Use the percentage at the end, to check your answer is not absurd. Do not start there.
The way to actually get a number
Work backwards from the customer, in four steps. Twenty minutes with a notebook.
Step one. What is a customer worth in gross profit? Not revenue. Revenue minus the cost of delivering the thing. If you have repeat business, take the first year rather than a lifetime, because a lifetime figure invites optimism.
Step two. What share of that will you spend to get one? Between 15% and 30% for most service businesses. Higher if you are growing deliberately and can fund the gap, lower if cash is tight.
Step three. How many enquiries does it take to close one? Your own number, from your own records, not an industry average.
Step four. Multiply out. Allowable cost per customer divided by enquiries per customer gives your allowable cost per lead. Then multiply that by the number of customers you want a month.
Worked through, for a dental clinic in Koramangala.
A patient is worth about Rs 12,000 in gross profit across the first year. You will spend 25% to acquire one, so Rs 3,000. You close one in four enquiries, so your allowable cost per lead is Rs 750. You want ten new patients a month, so you need forty enquiries, so your budget is Rs 30,000.
Now check it against the floor. Forty conversions a month clears thirty comfortably. This campaign can work.
Run the same arithmetic for a business where the allowable cost per lead comes out at Rs 200 and the target is four customers a month, and it produces a budget of Rs 3,200 for sixteen enquiries. That campaign cannot work on paid search, and the honest answer is not to spend the Rs 3,200 anyway.
What the ranges actually look like in India
For context rather than as a target, and these are the bands I see rather than a survey.
Below Rs 25,000 a month you are choosing one channel and doing it properly, or you are doing the unpaid work instead. Split across two platforms at this level and neither reaches its floor.
Rs 25,000 to Rs 75,000 a month funds one paid channel at a workable volume plus some content, for a local service business in a single city.
Rs 75,000 to Rs 2,50,000 funds two channels, proper creative production and somebody competent running it.
Above that you are into multi-city or multi-channel and the constraint stops being budget and starts being whether anybody is managing it.
The Indian agencies publishing 2026 benchmarks put Google Search clicks around Rs 24 on average, with ordinary categories between Rs 8 and Rs 25, and Meta lead costs commonly between Rs 150 and Rs 400 for service businesses. Plug your own category into the arithmetic above rather than taking those as your numbers. I went through which of the two platforms your business should start on separately, and that decision comes before this one.
If you are under the floor
This is most people reading this, so it deserves a real answer rather than a suggestion to spend more.
Pick one channel. All of it, one place. Two half funded campaigns is the single most common way small budgets get wasted in India, and it feels like prudence while being the opposite.
Then narrow until you clear the floor. One city, not four. One service line, not the whole menu. A Rs 20,000 budget aimed at one procedure in one neighbourhood can absolutely reach thirty conversions. The same money aimed at everything you do reaches nothing.
And if the arithmetic still does not work, stop and go do the things that cost time instead of money. Google Business Profile filled in properly and posted to weekly. Reviews asked for after every job. WhatsApp replies inside ten minutes instead of the next morning. Those move numbers for local Indian businesses more reliably than a Rs 10,000 ad budget ever has, and they cost nothing but attention.
Two things about the money itself
Advertising services in India carry 18% GST. If you are registered, that is generally recoverable as input credit, which means the sticker price and the real cost are not the same number. Confirm your own position with your CA, and read the invoice rather than the dashboard. There is more on that in what digital marketing actually costs in India.
Second, separate the media budget from the management fee in your own planning. A Rs 50,000 total that turns out to be Rs 30,000 of media and Rs 20,000 of fee is a campaign with a Rs 30,000 floor to clear, not a Rs 50,000 one. People routinely calculate this wrong and then wonder why the volume never arrived.
Before lunch
Get your last twelve months of enquiries and closed customers. Work out the gross profit on one customer and how many enquiries it took. That is two numbers and they decide everything above.
Multiply your current cost per lead by thirty. If your monthly paid budget is below that figure, you have a structural problem rather than a performance problem, and the fix is narrowing the campaign rather than changing the copy.
My notes on performance marketing cover how the rest of it fits together. If you’d like to sanity check your own arithmetic, drop me a line on email, WhatsApp or LinkedIn and we can have a quick chat. I’m contracted full time so this isn’t a pitch.