Ek din. Idea dekhi ad live samma.
The ad I am describing is a client's, it is running now, and it was not a rushed job. Here is the method in three steps — then the part nobody here writes about: what actually caps a Nepali business's ad spend, and why it is not the marketing.
Facebook ad chalauna kati samaya lagcha?
One day, if the decisions are already made. The same job used to take me two to three weeks minimum, and on a bigger campaign two to three months. The time was never the setup screen — Ads Manager is clicking. It goes into three decisions: the offer, the campaign type, and where the creative comes from. Each used to be a week of my own work. Each is now a conversation, but only if you know what a good answer looks like.
Ad chalaunu bhanda pahila ke ready hunu parcha?
The offer. Not the creative, not the budget. Read $100M Offers by Alex Hormozi, then Dan Kennedy's Magnetic Marketing — a small book that returns far more than it costs.
If you have no time, shortcut it with AI, with a warning attached. What you get will not match reading the book yourself, and as of the day that video went up, AI does not beat a human expert at this. Understand it first, then delegate it.
The prompt, exactly as typed on camera:
I want to start a clothing brand in Nepal for child
>
use Alex Hormozi and Dan Kennedy way of giving offer and come up with offer that I can run in Meta ads
The first line is easy — anyone can describe their business. The second line is the part people cannot write, because it takes knowing whose method to apply and where the output has to run. That is the skill; the typing is not.
Pahilo answer lai cross-check garnus
Do not implement what comes back. On that first prompt it pushed back at me — it said Hormozi's framework is service-based and would not carry a product offer well.
Let it think, too: the better the model, the longer it sits there. Then send it back to check its work against the market:
Can you do a research on best offer on clothing brand for child? What works and what don't and what others are doing?
Kun AI tool use garne?
Whichever one you can pay for — but that pushback is why I use Claude and Claude Code. Ask ChatGPT the same thing and it tends to agree, then flatter you: you know Dan Kennedy and Alex Hormozi, most business owners in Nepal don't even know who they are. Flattery is not information, and you are about to spend money on this output. Codex does the same job, and the Chinese models will run these prompts if cost is the constraint. Kura AI tool ko haina, method ko ho.
Meta ads ma kun campaign type chalaune — lead ki awareness?
For my client it was awareness, and that one change is what made the ad cheap. I had been running lead campaigns only, always, and they were expensive. The advice was to stop — my demographic was not an elite client base, and for that audience awareness or engagement brings better response for less money.
Then the setup prompt:
guide me step by step on setting up Meta Ads. What type of campaign I should run? How much budget and tell me everything step by step.
Back comes campaign type, budget, pixel and conversion API setup, in order. If you do not know what a pixel is, ask that directly.
I ran it that way. The ad was a hiring ad — a client needed trainers. Around $8 of spend, and somewhere around 200 to 300 people applied to work at that client's company. I hedged that on camera; the hedge stays here.
Read what that is and is not: applications on a hiring ad. Not reach, not leads, not sales. A hiring ad is the cheapest ad most Nepali businesses will ever run, because you are asking for time and a form, not money.
Applications are not the outcome, though — hires are. From that same campaign the client hired 3 people, and the site took 1,831 visits. So the honest chain is $8 in, 200 to 300 applications, 3 people actually working. Every stage of that funnel drops hard, which is the point: a number that sounds enormous at the top is a small, real number by the time it matters, and you should read anyone's ad results that way — including mine.
Following its suggestions on Facebook has not lost me money so far. That is a record, not a guarantee. Test on a small budget.
Nepal bata Facebook ad ma kati kharcha garna sakincha?
Directly, up to $500 — the limit on the dollar card the Nepal government provides me, and the real ceiling on this whole method. Above it, you are either asking friends abroad to pay or going through a marketing agency.
The agency I work with — Param Digitech in Kathmandu, +977 9702990030 — provides an ad account and proper bills for whatever is spent. How they arrange it on their end I do not know and do not ask. What matters to an owner is that the spend is documented — which matters the moment anyone looks at your accounts. With that in place you can run whatever budget you want.
Notice what this means. If your ad spend is capped, that is a payments problem, not a marketing problem. Owners diagnose it as the second one constantly — "Facebook ads don't work in Nepal" — when the constraint is a card limit and a missing ad account.
The same thing unlocks the automation. With Meta Marketing API access the AI can run the ad account itself — creating campaigns, not just advising on them. I have not done that, and the reason is not the technology. It is the card. I asked what would be required and one line came back: a Facebook Meta app with marketing API access. Your own card, your own account, or an ad account provider who hands you that key — any of the three, and campaign management becomes a conversation.
Ad ma price rakhne ki narakhne?
Put it in — at least in the case I ran, though the standard advice says the opposite. The two most powerful words in marketing are free and new, and both carry a cost people ignore. New means changing how you deliver the service, which shakes the business internally. Free brings people who were never going to buy — free ma khana ko lagi giddha nai aaidincha.
I asked how to solve that and got a genuinely creative answer: put the price in the ad.
State the price and most people do not come. The ones who do come knowing what it costs — and that person is a lead, not a customer. You still call, follow up and sell before they buy, and not every business can do that part. At the client's gym, reception does not follow up well and the manager is doing everything at once, so the ad does the filtering the front desk cannot.
Three reasons it holds here. Ads are cheap in Nepal. The price removes the people who will not buy, for free. And while the ad runs, everyone seeing it learns what the place costs — my price is not a bargaining price, and hearing it repeatedly builds appetite instead of resistance.
Change one variable and the strategy might change with it. But note where the idea came from: the AI produced it, and I only got there because I could name the problem. AI cha bhandaima tapai lai marketing aaudaina. It answered a good question; writing the question was the job.
Video editor bina ad ko creative kasari banaune?
From footage you already own. Last year I had photos taken of every trainer and staff member at the client's gym — one shoot, one videographer. Those photos now go to the AI, which edits them and places the banner and headline. The New Floor Trainer line on the creative running right now was placed that way, no editor involved.
Put the photos and videos in Google Drive and give the link once. One of the videos came from an Instagram link I pasted, which it downloaded itself.
Now the restriction: do not generate the footage. I rejected most of what was generated — generated video reads as generated, and people neither trust it nor rate it. I rejected another because it carried Nepali text I could not read properly. If I cannot read it, my customer cannot.
AI does basic editing, and basic is sometimes enough. Watch it as a viewer before it goes near production. If it does not hold up, get a videographer and a real editor — footage shot properly once keeps working, which is why last year's photos are still feeding this.
Toh kaha bata suru garne?
Offer, campaign setup, creative — in that order. Most owners start at the creative because it is the visible part, and that is why the ad underperforms: a good creative on a weak offer only buys a cheaper way to be ignored.
Tapai ko last ad ma campaign objective ke thiyo — lead, ki awareness, ki engagement? And did you choose it, or did it arrive as the default? Most owners here have never made that choice deliberately, and for me it was the single change that moved the cost per result.
Then the small thing, today. Take the first prompt above, replace clothing brand in Nepal for child with what you actually sell, keep the second line as it is, and read what comes back. Do not run anything — just notice whether it argues with you or agrees with you. That tells you what you are working with before you spend a rupee.
Full walkthrough, with the ad account and the rejected creatives on screen: Nepal ma Facebook Ad Kasari Chalaune?
Ad ma risk huncha. Sano budget ma test garnus. Malai aile samma ghata bhaeko chaina — tara tapai lai bhayo bhane ma jimmewar chaina. Buddhi bibek prayog garera garnus.