Most advice about selling online in Vietnam starts with a platform checklist. That is the wrong end of the problem. A small shop rarely dies because it chose Shopee over TikTok Shop. It dies because the second order ships late, because the stock figure on the marketplace disagrees with the shelf, or because nobody answers a message on a Sunday evening. Platforms buy attention. Operations are what keep the money.
What follows is the version of this advice I would give a friend opening a small store this year — organised around the promises a shop has to keep, not around the list of apps it should install.
Start with the promise, then pick the channel
Before comparing fees, write down what you can reliably deliver: how fast you can pack, which items you can restock within a week, and how many conversations you can answer in a day. A channel that brings impulse buyers is a bad fit for a product you can only restock slowly, and a marketplace built on price competition is a bad fit for something you assembled by hand.
That single test usually sorts the options on its own:
- Marketplaces with huge ready-to-buy traffic suit standard, high-volume items. You accept commission and price pressure in exchange for reach.
- Short-video commerce suits anything that can be demonstrated in a fifteen-second clip and bought on impulse.
- Facebook and Zalo carry community, local targeting and the private conversation where a hesitant buyer finally commits.
- Your own site is the only channel you actually own. It protects margin and holds the catalogue you send people to when they ask what else you sell.
Most healthy small shops end up running two of these, not four. Spreading a thin team across every platform produces slow replies everywhere.
Payment is a trust decision, not a checkout setting
In this market the payment method is read as a signal about the seller. A buyer who has never heard of you will look for the option that lets them keep control of the money until the parcel is in their hands. Cash on delivery still does that job, and refusing it costs orders from first-time customers.
Digital methods matter for the second and third order, when trust exists and speed matters more. Offer a transfer or e-wallet option, print the account details on the confirmation message, and state clearly what happens if a transfer arrives after the parcel ships. Ambiguity here turns into support tickets and refunds.
Delivery is part of the product
Two things ruin an otherwise good shop: a processing time that was never realistic, and a delivery promise made at checkout that the courier cannot meet. Both show up later as cancellations and bad ratings, which marketplaces punish by burying the listing.
Decide your cut-off hour, stick to it, and describe it in the listing rather than in a message nobody reads. Send one update when the parcel is packed and another when it leaves. Those two messages remove most "where is my order" conversations before they start.
Returns deserve the same treatment. A return policy that is easy to find and easy to use is a sales asset, not a cost: buyers who know they can send something back are more willing to try a shop they have not used before.
Content that convinces is not the same as content that looks expensive
A phone, daylight and a clean background beat a rented studio for most small catalogues. What buyers respond to now is evidence that the product exists in the real world: it in hand, in use, at the size it actually is.
Three kinds of material do most of the work:
- A short demonstration of the product solving the problem it was bought for.
- An unboxing that shows exactly what arrives, including the boring parts.
- Something from behind the counter — how it is packed, where it comes from, who answers the messages.
Educational clips outperform pure promotion because they get watched to the end. If you sell skincare, explain skin types. If you sell tools, show the job finishing faster.
Stock records are the quiet failure point
Overselling happens when the same unit is sold twice — once on a marketplace and once in a direct message. The fix is not more discipline, it is one stock number that every channel reads. A simple inventory tool that pushes the count to each platform as soon as an order lands prevents the cancellations that marketplaces treat as seller faults.
The same logic applies to the files behind the shop: product photos, supplier invoices, order and return records, monthly sales exports. They are the evidence you need when a buyer disputes a claim or a platform asks for proof. Keep them somewhere other than the one phone and one laptop the business runs on — a wiped device should not take the shop's history with it. Folder sync with version history and automatic upload from the phone is the practical minimum; when comparing storage plans, check whether the price is genuinely one-time or renews higher later, and whether the capacity is sized for video and raw images rather than documents.
Replies are a sales channel
Speed of reply separates shops that look alive from shops that look abandoned. A question waiting a couple of hours on Zalo usually goes to whoever answers in minutes.
Save the answers to the questions you get every day — price, stock, delivery time — as quick replies so the first response is instant and human. Then use the conversation for the part automation cannot do: sizing advice, substitutions, the extra item the buyer did not know you stocked. Follow up once on a cart that was filled and abandoned; more than once reads as pressure.
The costs people forget to budget
Commission on every marketplace sale, ad spend to keep a listing visible, packaging good enough to survive a courier, and small monthly fees for inventory and customer tools. None of these is large alone. Together they decide whether the margin you calculated on paper survives contact with a real month. Price with all four included, not just the cost of goods.
Mistakes that repeat across shops
- Guessing what sells instead of reading the marketplace analytics that are already free.
- Looking like a different brand on each channel, which quietly erodes trust.
- Treating processing time as an internal detail and then missing it publicly.
- Buying traffic before the reply speed and stock accuracy can handle the orders it brings.
Selling online here is not a technology problem. It is a reliability problem with a technology layer on top. Pick the channels that fit what you can actually deliver, make the payment and delivery promises explicit, keep one honest stock number, and answer fast. A small shop that does those four things can hold its own against competitors many times its size.
