What Is Live Commerce? A Practical Guide for Thai Sellers

Live commerce is selling through a live video broadcast where the shopping cart and the payment flow sit inside the stream itself. Viewers buy from the same screen they are watching, without searching for the product somewhere else and without sending a chat message to ask the price. What separates it from ordinary live selling is that the order, the stock deduction and the payment all happen inside the same system the audience is already in, which shortens the distance between interest and payment to a few seconds. This article covers how live commerce works, how many people and how much kit a broadcast needs, how to price during a live, which numbers to judge it by, and what makes a room go quiet with no orders at all.

What live commerce is, and how it differs from ordinary live selling

The style of live selling Thai shoppers have known for years works like this: open the camera, talk, and let customers type CF in the comments. The seller then collects comments line by line, totals them up, sends bank details, waits for slips, checks the slips, and only then packs the goods. It sells, and people still do it every day. The back office cost is high once viewership grows, and the faster the comments scroll, the more orders slip through. Payment outside any system also leaves no central record to check later when there is a dispute about a missing item or the wrong colour.

Live commerce moves that whole back office into the platform. The seller loads products into a list in advance, then pushes the item being discussed onto the screen. Viewers tap the product card and land on the app's checkout while the stream keeps playing in the corner, so the buyer never leaves the show. The measurable consequences are few and specific.

• The order is recorded the moment the viewer taps, with no need to go back through comments afterwards.
• Stock is deducted against real quantities, which cuts the cases where you oversell and have to cancel on a customer.
• Views, add-to-cart actions and payments land in one report, comparable against the same timeline.
• Payment and parcel tracking stay in the same app, so the customer is not switching between apps halfway through.

The content format differs too. Lives that sell tend to be built like a television shopping programme: a product order set in advance, planned slots for which item comes up when, and information repeated in cycles. The reason is simple. Viewers arrive throughout the broadcast, and someone who joins at minute twenty never heard the price given at minute three. Without repeating the price, the sizes and the ordering steps, that group leaves quietly without asking anything.

What live commerce does better than a static product page is answer doubts in real time. Several product categories are hard to sell from still photos: clothing where the drape matters, kitchenware where the real size next to a hand matters, food where the texture at the moment of cutting matters. Viewers can type a question about whether it fits tight or whether it is dishwasher safe and get an answer in seconds. That is the part that comes closest to replacing a physical shop.

Which platforms support live commerce in Thailand

Sellers here use four main channels, and each brings a different kind of audience. That difference, not the platform's total user count, is what should drive the choice.

TikTok

TikTok has an in-app shop that connects to the live stream. The seller adds products to the broadcast's list and pushes the item being discussed onto the screen for viewers to tap. Its strength is the recommendation system, which puts the live in front of people who have never followed the shop, so a room can grow quickly if the first thirty seconds are worth watching. The thing to watch is category rules, since some categories require extra documents or a licence before you can sell. Open the seller centre and read the restricted and prohibited category list before planning a month of broadcasts. If you plan to use paid media to bring people into the room, the campaign setup and creative formats are covered on the TikTok Ads page.

Shopee

Shopee Live sits in the same app as the shop and the cart customers already use. People who come in are usually in buying mode rather than browsing mode. Sellers can set live-only discount codes, free gifts and giveaway activities from the seller centre. The part that needs planning is getting people into the room, because discovery in a marketplace leans on follower notifications and the placement the app gives you, rather than being pushed out to new users at random.

Lazada

Lazada has an in-app live feature as well, structured much like Shopee's: tied to the shop and the existing payment system. It suits shops that already have a follower base and use lives to clear stock or launch new items in cycles. Time-limited coupons and telling followers the schedule in advance are the two things that keep early viewer numbers from dropping so far that the app stops surfacing the room.

Facebook

Facebook is still one of the biggest places Thai sellers go live, especially in closed groups and pages with regular customers. The advantage is familiarity and a conversational tone. The limit is that the cart is not wired in as directly as on a marketplace, so many shops use an external assistant tool that reads comments containing a product code and issues an order with a payment link into the customer's chat. That works well when product codes are short and repeated often, but someone still has to handle the cases where a customer mistypes.

If you have to pick one channel to start with, look at where your current customers already buy. A shop with reviews and accumulated ratings on a marketplace should start there, because the live inherits that existing credibility. A brand with no customer base that needs new people should start where the recommendation system works harder, then add a second channel once the team can run a broadcast without drama.

How much equipment and how many people a live needs

The kit that actually matters

The starting set does not need to be expensive. The thing that makes a live look bad is rarely the camera. It is sound and light. Viewers tolerate mid-range image quality, but they will not tolerate echo or a voice buried under background noise for more than two minutes. What is worth having from the first broadcast:

• A phone with a steady camera and focus that does not hunt, plus a second phone to watch the viewer-side screen.
• A tripod that locks firmly and holds the camera at eye level, not angled up from a table.
• One or two lights placed off to the front of the host so shadows do not fall across the product.
• A clip-on or wireless microphone separate from the phone's built-in mic. A room with bare walls needs fabric or a rug to cut the echo.
• Internet with a backup, such as fibre as the primary and a mobile SIM as the fallback. A dropped connection mid-live means starting the viewer count from zero.
• A table or backdrop that holds products in order, so nothing has to be hunted for off camera.

The step up, once broadcasts get frequent, is a separate camera with a capture card, studio software on a computer for price graphics and countdown timers, and a second monitor so the host can see comments without glancing at a phone. Buy those after the lives have proven they make money, not before.

The team and who does what

The work in one broadcast splits into five jobs, which does not mean five people.

• The host, who talks on camera, controls the pace and repeats prices and ordering steps in cycles.
• An assistant on or behind camera, who hands over products on cue, changes price cards and checks that what is shown matches what is loaded in the system.
• A comment moderator, who answers the repeat questions about shipping cost and delivery time and passes harder questions to the host.
• A back office operator, who loads products, pushes product cards, releases coupons and watches remaining stock during the broadcast.
• Packing and dispatch, which can start during the live once orders are already in the system.

The smallest team that genuinely works is two: a host, and one person covering both comments and the back office. Once orders per broadcast rise to the point where that second person cannot type fast enough, split off the back office operator as the third hire first, because pushing the wrong product card at the wrong moment costs more sales than answering a comment slowly.

How to price and run promotions during a live

Pricing in a live behaves differently from pricing on a product page, because viewers decide under time pressure and can see what other people are buying. A price structure that holds up to scrutiny usually runs in this order.

1. The normal price has to stay visible in the shop at all times. Do not inflate the shop price so the live discount looks bigger. Beyond the advertising rules, regular customers spot it very quickly.
2. An opener is a low-priced item in a genuinely limited quantity, there so newcomers can make a first tap. That first purchase is what keeps a person in the room.
3. The hero item is the profitable one you set out to push that session. It should get the most talk time and be repeated at least three times across the broadcast.
4. Bundles lift order value, for example by grouping items people already buy together and adding a shipping discount.
5. A closer is announced near the end, so viewers have a reason to stay and undecided people get one last chance to commit.

An illustrative example of the thinking: say a broadcast runs ninety minutes with twelve products. Split it into four blocks of twenty to twenty-five minutes, each with one lead item and two supporting items, and close each block by repeating all three prices and the ordering steps. Those figures are an example of a time structure, not an industry standard. A shop selling high-priced goods that need explaining might cover half as many items in the same time.

Coupons carry one trap worth knowing. Push hard live-only discounts too often and regular customers stop buying at the normal price and simply wait for the next broadcast. Total revenue may look like it grew, but margin per unit falls and weekday sales disappear. The safer route is gifts or privileges that do not touch the headline price, such as samples, faster dispatch, or a bundle that cannot be bought as separate items.

Which metrics to judge a live by

Total views is the best-looking number and the least informative. The table below is the set worth recording after every broadcast, with what each one means and what to do when it looks wrong. All of these can be read from the post-live report in each platform's seller centre without any extra tooling.

Record this set in the same file every time, along with three notes: who hosted, which item was the lead, and whether paid media was running. After about ten sessions, patterns you can act on start to appear, such as which broadcast times bring buyers rather than watchers, or which host converts best with which product type. Setting up that record is part of the ecommerce marketing work that lets a live feed into the rest of the channels.

How to use ads to drive a live to the right audience

Paid media around a live splits into three moments: before, during and after. Each has a different goal and should be measured differently.

• Before the live, short clips announce the time and the featured items. The goal is a follow or a reminder. The most efficient audiences are people who have watched the shop's clips and people who have bought before.
• During the live, use the ad formats that take people straight into the room. This budget only works while you are on air, so set it up in advance and switch it on as the show starts. Judge it by how many people entered and stayed past one minute, not by clicks.
• After the live, cut clips from that session and run them at people who watched but did not buy, and at people who added to cart and did not pay. That last group usually returns the most per baht spent.

The warning is not to open a large live-boost budget in the first few sessions if the room has no conversation in it yet. New people brought into a silent room leave within seconds, and that signal tells the recommendation system to stop surfacing the broadcast. The less damaging path is to build followers and a consistent schedule first, then add budget to the slots where you already know people show up.

Mistakes that stop a live from selling

In broadcasts that draw viewers but no orders, the problem is usually not the price. It is a mechanic that can be fixed in the next session.

1. Not loading the product into the system, or loading it later than you talk about it. Viewers want to tap while they want the item, and if there is no card to tap, the impulse is gone in seconds.
2. Not repeating prices and ordering steps. New people arrive constantly, so repeating every ten to fifteen minutes is basic practice rather than boring filler.
3. Falling behind on comments, especially on shipping cost and delivery time. Those two questions block payment more than any other, so keep canned replies ready and have the host say the answers out loud at intervals.
4. Stock that does not match, forcing cancellations after the sale. Shop ratings suffer and customers remember. Reserve a separate live allocation apart from the shop's main stock to prevent it.
5. Sound and light that fail, so viewers leave before hearing the content. This is fixed with inexpensive gear and should be fixed before any other spending.
6. Broadcasting irregularly, going live once and then disappearing for two weeks. Viewers cannot remember a schedule, and the recommendation system has too little history to surface the room.
7. A host who reads a script and never answers what people ask. A live like that offers nothing over a pre-recorded video the viewer could watch whenever they liked.

Live commerce and Thai buyer behaviour

Live commerce in Thailand runs on conditions that differ from some other markets. First, buying through conversation has been normal here for years, going back to selling in Facebook groups and messaging a shop to ask the price. Live selling is not a new behaviour that has to be taught from scratch. Second, paying by transfer and QR is an everyday habit, so the payment step is a smaller obstacle than it is in markets still built around cards.

What does remain an obstacle is shipping cost and delivery time, particularly for customers outside Bangkok. Shops that state shipping clearly at the start and hold the conditions steady through the broadcast tend to get better payment rates than shops that let customers discover the cost at checkout. Language matters too. Saying product names and sizes in the Thai wording people actually search for helps clips cut from the live get found later. Many shops ignore this because they treat a live as finished when it ends, when the replay can keep producing sales for days.

Last is timing. Weekday evenings and the early afternoon on weekends are slots many shops use, but the right answer for any one shop comes from its own data. Try three different slots, three sessions each, then compare sales per live hour across them. That is a more reliable answer than copying somebody else's schedule.

Live commerce FAQ

How many followers do you need to start live commerce

There is no required minimum, because platforms with a recommendation system can bring new people into a room even for a shop that just started. What matters more than follower count is a consistent broadcast schedule and the quality of the first thirty seconds. A shop going live twice a week without gaps usually sees results sooner than a shop that goes live once and stops.

How long should one live session run

The right length depends on how many products you carry and how long each one takes to explain, rather than any standard figure. The usable rule is to give each item enough time to demonstrate and answer questions, then leave room to repeat prices in cycles. If a session runs long enough that the host starts going in circles, cut the product count instead of stretching the time.

Do you need a studio, or can you broadcast from the shop

You can broadcast from the shop floor or the warehouse, and in many cases it works better because viewers see the real goods and the real setting. The conditions are controlling background noise and light, and having a corner where products sit in order and are easy to reach. If the shop is loud enough that the mic picks up conversations behind you, choose a closed corner or a quieter time of day.

Can high-priced products be sold through a live

Yes, but the format of the show has to change. Expensive products need more explanation time, a more detailed demonstration, and often close through a follow-up chat rather than an immediate tap to pay. The metric should shift from sales per hour to the number of qualified enquiries received and the close rate after the broadcast.

What can you do with the replay afterwards

Cut the demonstration segments and the better question-and-answer moments into short clips, then use them as feed content and as ad creative for people who watched but did not buy. Name the files and write the captions using the product wording people genuinely search for, so the clips can be found through in-platform search.

Live commerce pays off for shops that run it as a system rather than as an occasional event, because everything inside it can be measured and corrected one piece at a time. If you want a team to plan the run sheet, run the sessions and read the post-broadcast numbers, the service details are on the Relevant Audience live commerce page.