How to Sell Online in Nepal: Choosing Where to Sell Before You Build Anything

The short answer

To sell online in Nepal, pick the channel your customers are already in before you build anything. There are four routes: sell directly through Facebook and Instagram, list on a marketplace such as Daraz, build your own store, or combine them deliberately. Social selling gives immediate reach and no ownership. A marketplace gives buyers who already trust the checkout, and takes margin and the customer relationship for it. Your own store gives you the customer, the data and the asset, and no traffic on day one.

For most first-time Nepali sellers the honest sequence is: start on social, where the audience exists and testing costs almost nothing; add a marketplace listing if your product is a comparison item people search for by name; build your own store once you have repeat buyers you cannot reach without paying a platform. Underneath all four sits the constraint that shapes Nepali ecommerce more than any other — most buyers pay cash on delivery, so an order is not revenue until the parcel is handed over.

Why building the store first is backwards

A website is not a shop on a busy street. It is a shop in a field, and nobody walks past. A store converts demand you already have; it does not create it. So the first question is not what to build, but where the demand already exists. That assumes it exists at all, which is a question worth answering first if the product is new — how to test demand before you commit money covers the cheap ways to check, and they are all cheaper than a store nobody visits.

Route one — Facebook and Instagram

This is how most online selling in Nepal begins, and the people doing it are not making a mistake. A page, product photos, a comment section, orders arriving as direct messages. It works for structural reasons.

• The audience is already scrolling. You appear inside a habit rather than asking anyone to visit a new place.
• Testing costs close to nothing. You learn whether people want the product in two weeks without building anything.
• The format matches how Nepali buyers behave — they ask questions first, and a comment thread is where that happens.
• Paid reach is immediate and can be aimed at one city, so a Pokhara seller advertises only to Pokhara.

What it costs you

You own nothing, and every limitation follows. No customer list to export, so last year’s repeat buyer is unreachable unless you pay again. No data beyond what the platform shows. No search visibility, so when someone types your product into Google you do not exist. And your order history is a scroll.

The ceiling is not volume — plenty of Nepali sellers move real volume through Facebook alone. It is compounding. Social-only selling produces revenue and no accumulated asset, and two sellers doing the same numbers are in different businesses by year three if one built something owned.

Route two — marketplaces

Daraz is the dominant marketplace in Nepal, and listing there buys two hard things: traffic from people already in a buying mood, and a checkout the buyer already trusts. The second matters more than sellers expect. Someone who would hesitate to pay an unknown page orders from a marketplace without thinking, because the trust attaches to the marketplace, not to you.

• Margin. Commission, fulfillment terms and promotions come out of your price, and they change — check current seller terms directly rather than taking numbers from an article.
• The customer relationship. The buyer belongs to the marketplace, and you generally cannot market to them later.
• Price comparison. Your listing sits beside everyone selling the same item, sorted in ways you do not control.
• Rule dependence. Ranking and promotion mechanics belong to the platform, and a change to them changes your revenue with no notice.

The honest test is product type. If people search for your product by name and compare on price and delivery — a phone case, a kitchen appliance, a standard consumable — refusing to list is refusing free traffic. If your product has a brand or a reason to be chosen that is not price, a marketplace flattens the exact thing you want to be paid for.

Route three — your own store

Your own store is the only route where you own the outcome: the customer list, the order history, the analytics, pages that can rank in search, and the ability to change how you sell without asking permission.

The trade is blunt. You are buying an asset that compounds and paying for it in demand generation. Traffic on launch day is zero, and every visitor for months arrives because you paid for them or did the slow work of search visibility. That is why so many Nepali stores are built, launched and quietly abandoned inside a year.

What makes it worth doing is that the costs move in opposite directions. Paid reach costs the same in month thirty as in month one; a store with returning buyers and ranking pages costs less each year. The build itself is covered in the companion post on starting an ecommerce business in Nepal, which picks up where this decision ends.

Route four — the hybrid, and how to sequence it

Almost every Nepali seller still trading in three years runs more than one channel, because they do different jobs. Social finds new buyers. The marketplace catches people already searching. The store holds the relationship and the repeat business. Running all three from a standing start is how small teams drown, so sequence matters more than combination.

A workable order: prove the product on social until orders are consistent. Add the marketplace if your product is a comparison item. Build the store once you have past buyers worth bringing somewhere you own, then move repeat purchasing onto it.

Cash on delivery is the constraint everything bends around

Most advice about selling online is written for markets where the buyer pays at checkout. In Nepal, cash on delivery is the dominant payment behavior, and it reshapes the business in ways no imported playbook accounts for.

What it gives you is real. A buyer who has never heard of you, in a market where being cheated online is a live fear, will order because paying later costs nothing if you never show up. It is a trust substitute, and the most effective one a new seller has. Removing it before you have a brand removes most of your orders too.

What it takes from you

• Refused deliveries. The customer changed their mind or is unreachable, and you pay to ship it out and back.
• Working capital locked up. You funded the stock and the courier, and the cash sits in transit until it settles.
• Casual ordering — the predictable behavior of a system with no commitment at the point of sale.
• An order number that is not revenue, which makes the business look larger than the bank account agrees with.

So your reporting needs two columns from month one — orders placed, and orders delivered and paid — broken out by product, city and traffic source. The gap is rarely even. Usually one product or one channel produces most of the refusals, and knowing which lets you fix it rather than abandoning cash on delivery.

Payment and checkout as a conversion problem

Payment in Nepal is usually discussed as an integration question. It is more useful as a conversion question, because checkout is where the largest single drop in any store happens, and it is the part sellers examine least.

Digital wallets and bank transfer are both common in Nepal alongside cash on delivery, and a prepaid option matters beyond cash flow: a buyer willing to pay in advance already trusts you, and giving them no way to do it wastes the signal. What I would not do is take a "best gateway" verdict from any article, including this one. Fees and terms change — compare them directly with each provider.

1. Every checkout step as its own event. A single conversion event says the checkout is losing people; step events say where.
2. Payment method recorded on the order, so you can compare the delivered-and-paid rate of prepaid orders against cash-on-delivery ones.
3. Failed and abandoned payment attempts. An error nobody retried looks identical to a customer who left, and the fixes differ.
4. Mobile separately from desktop. Nepali buying happens overwhelmingly on phones, and a small-screen fault is invisible in a blended report.

The principle is the one that governs ad measurement too: if a number cannot be traced to money that arrived, it is not a result.

Trust is the actual conversion lever

In markets with mature buyer protection, the platform absorbs the risk of a bad seller and the buyer clicks assuming someone will make them whole. That assumption is much weaker in Nepal, so the buyer runs the risk assessment themselves, in the seconds before deciding. Every trust signal is an input to it, and every absent one too.

• Real photographs of your actual product, taken by you. Stock images read as a warning, because they are what a fake page uses.
• A phone number a person answers. Not a form. An unanswered number is worse than none.
• "Based in Pokhara" with a real address. A named location beats design polish — it says there is somewhere to go if this goes wrong.
• A return policy in one plain paragraph, visible before checkout rather than buried.
• Visible responsiveness, plus your own name and face on the store. A seller willing to be identified is a credible one.

When your sales channel is a chat thread

WhatsApp and Viber are not support channels in Nepal. They are sales channels, often the primary one — the whole conversation from first question to confirmed order happens in chat, and the customer expects it to. Fighting that loses. What you can do is stop letting the thread be the only place the order exists.

• Nobody else can serve the customer. The context lives on one phone, so when that person is away the customer starts over.
• You cannot count anything — how many enquiries became orders, or which product converts worst.
• Follow-up depends on memory. Orders die because a message was read and never answered.
• Delivery details get transcribed by hand, which is where wrong addresses and refused deliveries come from.

The fix is not to stop selling in chat. It is to make chat the conversation and something else the record: a shared inbox rather than a personal phone, and every confirmed order written into one sheet with customer, product, address, amount and status.

The Nepal platform constraint nobody mentions

Here is a fact that changes the calculus and that almost nothing written about Nepali ecommerce mentions: Google Merchant Center, Shopping ads and free product listings are not available for Nepal. Nepal does not appear on Google’s published list of supported countries and regions for Merchant Center, and neither do Bangladesh, Pakistan or Sri Lanka. India is the only South Asian country on that list.

The consequence is structural. In most markets the product feed is the backbone of ecommerce demand — your products appear with a photo and a price in front of someone who typed exactly what they want. That route is closed to a Nepal-based merchant, and retail-goal Performance Max closes with it.

Demand has to be won elsewhere, so weight shifts onto search text ads, social advertising, marketplaces and content that ranks. That is a real argument for marketplaces on standard products: it is where feed-style discovery still happens in Nepal.

Where demand comes from once you have chosen

Channel choice decides where you sell. Demand generation is the separate question of how anyone arrives. Start with digital marketing in Nepal for how the channels fit together, then the Google Ads guide for Nepal for search, including the platform and billing limits specific to advertising from Nepal, and Facebook ads in Nepal for paid social, which is where most Nepali sellers spend first.

One sequencing note: do not turn on paid traffic until your delivered-and-paid rate is known. Advertising into a funnel that loses orders at the door scales a leak, not a business.

How to decide, in order

Work through these in sequence. The first one that gives a clear answer is your answer — do not skip ahead because a later option is more interesting.

1. Do you know people want this product at this price? If not, sell it on Facebook or Instagram this month. Building will not answer that.
2. Do people search for your product by name and compare on price and delivery? If yes, list on the marketplace. If it needs explaining, skip it.
3. Do you have repeat buyers you cannot reach without paying a platform? If yes, build your store, and let its first job be holding those relationships.
4. Can you fund demand generation monthly for six months? If not, do not build yet. A store with no traffic budget is a bill, not an asset.
5. Can you handle the cash-on-delivery operation at the volume you want? If not, fix that first — each new channel multiplies the same problem.
6. Only now choose the build. Platform, catalog and logistics are downstream of the answers above.

Where this goes wrong

• Building the store first, then discovering demand generation is a second project with a budget nobody planned.
• Treating a marketplace listing as a strategy. It is a rented channel, with no defense against a terms change.
• Scaling ad spend against orders placed. The number rises, the bank balance does not, and nobody sees why until the courier statements are reconciled.
• Discounting to fix a trust problem. If people doubt you will deliver, a lower price makes you look less credible, not more.
• Running three channels with one person, until stock, prices and response times drift apart.

When not to sell online at all

If your margin cannot absorb delivery cost plus a meaningful rate of refused deliveries, online selling makes each sale worse rather than better. If your product is genuinely local and walk-in demand is not saturated, the effort belongs there. If you cannot answer messages within a few hours, social selling will fail, because response speed is as much the product as the item. And if the reason you want a store is that a competitor has one, that is not a reason.

Registration, PAN and VAT

Selling online at scale in Nepal raises company registration, PAN and VAT questions, and import rules if you source from abroad. Those are real constraints, and they are outside the scope of this article and outside my expertise — I am a growth consultant, not a lawyer or an accountant. Take them to the Office of the Company Registrar and the Inland Revenue Department, and involve a Nepali chartered accountant early.

Where to start this week

If you sell nothing online today, put the product in front of people on Facebook or Instagram and record every buyer’s name and number from the first order. If you already sell through chat, get your orders into one sheet and calculate your delivered-and-paid rate. If you have a store with no traffic, the store is not the problem. I work on this through Arcetis, the growth systems practice I run from Pokhara. The pattern is consistent: the channel was chosen by accident and the store built too early.

Related reading