How much should a small business in Nepal spend on Facebook ads?

Most small businesses in Nepal should start at NPR 10,000 to 15,000 per month in ad spend. Below that, Facebook does not gather enough data to find your buyers, and you end up paying for guesses rather than results.

That number surprises people. The common assumption is that NPR 2,000 or 3,000 is a reasonable test. It is not, and understanding why saves you from concluding that "Facebook ads don't work in Nepal" when what actually happened is that you never gave the system enough to work with.

Why NPR 3,000 a month does not work

Facebook's ad system learns who buys from you by watching what happens after people see your ad. It needs a minimum volume of results before it stops guessing and starts targeting properly. Meta's own guidance is roughly 50 conversions per week per ad set before the system exits its learning phase.

At NPR 3,000 a month, you are spending about NPR 100 a day. In Nepal, a reasonable cost per lead for a café, clinic, or retail business sits somewhere between NPR 80 and NPR 400 depending on the industry and how well the ad is built. So NPR 100 a day buys you somewhere between zero and one lead. The system never learns. Your cost per lead stays high, and you conclude the platform is broken.

At NPR 15,000 a month, roughly NPR 500 a day, you generate enough activity for the targeting to actually improve over the first few weeks. That improvement is the entire value of the platform.

What different budgets actually buy you

NPR 5,000 per month. Enough to boost posts and stay visible to people who already follow you. Treat this as brand presence, not customer acquisition. Do not expect leads and do not judge it on leads.

NPR 10,000 to 15,000 per month. The realistic entry point for lead generation. One campaign, one clear offer, two or three creatives being tested against each other. Expect the first month to be learning and the second month to be where the numbers start making sense.

NPR 25,000 to 40,000 per month. Now you can run two campaigns at once, usually one for people who have never heard of you and one for people who visited your site and left. The second one is almost always cheaper per lead, and most businesses in Nepal skip it entirely.

NPR 50,000 and above. Full funnel. Awareness, consideration, retargeting, and enough volume to test seriously rather than guess.

Should the budget be a percentage of revenue?

A common rule is 5 to 10 percent of monthly revenue on marketing. It is a reasonable starting point, but it breaks in two situations that are common in Nepal.

If you are a new business with low revenue, the percentage gives you a number too small to work. Spend what you can afford to lose for three months instead, and treat it as buying information about your market.

If you have high revenue but a long sales cycle, like real estate or education consultancy, the percentage overstates what monthly ad spend should be, because a single conversion may take months to close.

Work backwards from a target instead. If you want ten new customers a month and your realistic cost per lead is NPR 300, with one in three leads converting, you need thirty leads, so roughly NPR 9,000 in spend. That calculation is more useful than any percentage.

What does a lead cost in Nepal?

It varies more by industry than most people expect. Broadly, in Kathmandu:

Restaurants and cafés tend to sit at the lower end, often NPR 50 to 150 per lead, because the decision is small and immediate. Retail and ecommerce fall somewhere in the middle. Clinics, dental practices, and education consultancies run higher, frequently NPR 250 to 600, because the customer is committing to something significant and takes longer to decide.

The number that matters is not cost per lead in isolation. It is cost per lead compared to what a customer is worth to you. A dental clinic paying NPR 500 for a lead that becomes a NPR 40,000 treatment is doing well. A café paying NPR 50 for a lead worth NPR 400 needs to think harder.

Does the management fee come out of the ad budget?

No, and any agency that blends the two is hiding something. Your ad spend goes directly to Meta. The management fee is what you pay someone to build, run, and optimise the campaigns.

If someone quotes you "NPR 10,000 all in," ask what portion is spend and what portion is fee. You will usually find that the actual money reaching Facebook is too small to produce results, and you will be paying for a campaign that was never given a chance.

At Narrative Machine we quote them separately, always. Ad spend is yours and goes straight to the platform. Our fee is for the work.

What to do before you spend anything

Three things, and skipping them is why most Nepali businesses waste their first month.

Know what a customer is worth to you. Without that number you cannot tell whether a NPR 300 lead is cheap or expensive.

Have somewhere for the ad to send people. A WhatsApp number that gets answered, or a page that explains the offer. Sending paid traffic to a Facebook page with no recent posts wastes the click you just bought.

Decide what counts as success before you start. Leads, calls, bookings, walk-ins. Pick one. If you decide after the fact, you will move the goalposts to whatever number looks best, and you will learn nothing.

The short answer

Start at NPR 10,000 to 15,000 a month in ad spend, run it for at least three months, and measure leads rather than likes. If your budget is genuinely below that, put it into organic content and Google Business Profile first, and come back to paid ads when you can fund them properly.

Spending NPR 3,000 a month on ads is not a small version of a real campaign. It is a different thing, and it does not work.

Narrative Machine is a digital growth agency in Kathmandu. We report on customers and sales, not clicks and impressions. If you want to know where your business currently stands before spending anything, our SEO Audit is NPR 5,000.