Selling Only Through Chat With No GP to Pay: Is a Shopping Cart Still Worth It?

Picture a small corner shop at the mouth of a soi. Nothing on the shelves has a price tag. If you want something, you walk up to the counter and ask the auntie sitting behind it. "Auntie, how much is this bottle of fish sauce?" "Do you still have the big bag of washing powder?" She knows every answer by heart, remembers every price, and asks after your kids while she's at it.

When customers come in one or two at a time, a shop like that feels wonderful. But on the day five or six people walk in together, a queue forms at the counter. Somebody who can't wait puts the item back on the shelf and quietly walks out. The auntie hasn't done anything wrong. It's just that every question in the shop has to go through one person.

A shop that sells only through LINE or a Facebook Page looks a lot like this. The chat window is the counter, the admin is the auntie, and "How much is this?", "Is this colour still available?" and "How much is shipping upcountry?" are the questions she answers on repeat all day.

I've written before about how much monthly commission you need to pay before your own website becomes the cheaper option. That formula works well for shops selling on Shopee, Lazada or food delivery apps. But if your shop sells purely through chat, you'll get zero when you plug in your numbers, because there's no GP to save in the first place.

Does that mean a shop like this never needs to think about a shopping cart? Not necessarily. A chat-based shop has real costs too. They just don't appear on a fee statement. They hide in the working hours of whoever is answering the chats. This article walks you through counting that cost with your own shop's numbers, so you can decide whether it's time to put price tags on the shelves.

Count One Day of Chats First

Pick an ordinary day that isn't a promotion day. Open every chat that came in that day, and sort your customers' messages into four piles.

• Pile one: questions with the same answer every time. Price, sizes available, colours left, shipping cost, the bank account to transfer to.

• Pile two: questions that genuinely need a person. Help choosing a gift, whether a product suits a certain skin type, a problem with an item that arrived.

• Pile three: chats that went quiet. A customer asked the price and never replied, or messaged at 1am and the shop only saw it the next morning.

• Pile four: orders that had to be retyped. Name, address and transfer amount that the admin copied out of the chat into somewhere else.

Once you've counted, the overall picture is enough. If pile one is clearly bigger than pile two, your admin is spending most of the day on work a price tag could do. If pile two is bigger, your customers genuinely want advice, chat is the heart of your shop, and a cart may not be the first thing to build.

Piles three and four tell you whether your counter has started to form a queue.

If your shop takes chats from both LINE OA and a Facebook Page inbox, count both. Many shops only count the inbox they check most often, and the number comes out smaller than reality. If you have more than one admin, have each person count their own chats and add them up at the end of the day. You don't need any software for this. One sheet of paper and a tally mark per message is plenty.

Where the Cost of a Chat-Based Shop Hides

A shop that pays GP sees its cost clearly every month, because it's deducted from sales as a line on a report. A shop that sells through chat has no report like that. The cost is spread across three places nobody writes down.

Time Spent on the Same Questions Over and Over

Say one admin answers dozens of pile-one messages a day. Each one takes almost no time. Type the price, send a photo, send the account number. But added up across a whole day, it becomes hours, and those hours have a real wage behind them.

I'm not going to give you a salary figure for admins. Every shop pays differently, and an average figure borrowed from somewhere else is more likely to push you toward the wrong decision than having no number at all. Take your admin's actual wage, divide it into an hourly rate, and multiply by the hours spent on pile one. The number you get is the "GP" of a chat-based shop: a cost you pay every month without ever seeing an invoice.

Customers Who Message After Closing Time

Chats accept messages all night, but the person answering them goes home. A customer browsing at 10pm asks a price, can't wait for the reply, and sometimes buys from another shop where they could order on the spot.

Pile three is the shadow of those customers. It doesn't tell you how many sales you lost, because there's no way to know whether the people who went quiet would really have bought. But it does tell you how often someone was standing at your counter when nobody was sitting behind it.

The Steps After a Customer Says Yes

In a chat-based shop, "I'll take this one" isn't the end of the job. The admin has to total the order, send the account number, wait for the transfer slip, ask for the address, type the address onto a shipping label, and in some shops copy everything again into a notebook or an Excel file.

That back-office side is a sizeable topic of its own. If your order records are starting to scatter across so many places you can't find them, I've covered it separately in the warning signs that your business has outgrown Excel and LINE. This article stays on the shopfront side, the moment when a customer is deciding to buy.

The Numbers That Support This Picture, and the Numbers That Push Back

Chat-based shops are far from alone. ETDA's Thailand Internet User Behavior survey for 2022 asked online sellers which channels they sell through (multiple answers allowed). Facebook came first at 66.76%, followed by e-Marketplaces at 55.18% and LINE at 32.05%. Sellers using their own Website or App came in at 26.67%.

On the buyer side of the same report, 39.70% said they had bought through a Website or App. These two figures come from different question groups, so you can't subtract one from the other. What they do show is that buying through a web page is already familiar to a large number of Thai shoppers. It isn't something new your customers would have to learn.

There's another figure in the same report that chat-based shops should know. 67.32% of online buyers pay by transfer through a banking app, and 58.49% have used cash on delivery. If you ever add a cart, its payment step has to support the transfers your customers are used to. A cart that only takes credit cards tells your regular chat customers they need to change their habits to buy from you.

Trust is the other side of this. ETDA's 1212 complaint centre statistics for 2025 recorded 14,238 complaints about online buying and selling, 36.40% of all complaints received, and the top category for three years running. So a customer meeting your shop for the first time has a real reason to hesitate before transferring money to a personal account number sent in a chat. A product page with a clear price and an orderly checkout can ease some of that hesitation, though it won't remove it entirely.

Now for the number that pushes back against everything above, and I think it's the most important one in this section. Baymard Institute's compilation of 50 studies (updated 22 September 2025) puts the average rate of shoppers who add items to a cart and then don't pay at 70.22%. Among US shoppers, the top fixable reason was extra costs appearing at checkout, at 40%. These figures come mainly from Western markets. I couldn't find an equivalent benchmark for Thailand.

In other words, having a cart doesn't mean everyone will pay. In a chat, the admin can answer that last doubt instantly. In a cart, nobody is there. A shop that closes sales because its admin is a great talker could lose that strength by moving everything into a cart. So I don't recommend shutting the chat and making customers order only through the cart. What suits a chat-based shop better is letting the cart take pile one off the admin's hands, while the chat keeps looking after pile two just as before.

"But Our Customers Love Chatting With Our Admin"

That's true. I'm not going to argue with it.

Plenty of chat-based shops sell well precisely because of how friendly the person answering is. Customers remember the admin's name, say hello before ordering, and some come back again and again because they enjoy talking to that particular shop. A relationship like that is hard to build and worth more than most people realise.

But look at pile one again. Most customers who ask "How much is this?" don't want a conversation. They want a price. If a price tag can tell them at 10pm, they don't have to wait for the auntie to wake up in the morning. Customers who really do want to chat can still message you, and your admin will have more time for them, because they won't be typing the bank account number for the fiftieth time that day.

The corner shop that puts up price tags still has the auntie sitting behind the counter. The tags just free her up to talk about things that matter more than the price of fish sauce.

Why the Longer You Sell Through Chat, the More Scattered Your Records Become

Every order closed in a chat leaves its trail inside that one customer's chat thread. If one day you want to know which customers reorder most often, which product sold best last month, or who bought this colour and might like the new version, the answers all exist. They're just spread across hundreds of chat threads, and you'd have to open them one by one to piece them together.

A shop that has sold through chat for three months can still trace things back. For a shop that has sold that way for three years, it's close to impossible. The longer you wait, the more of your history slips out of reach. Orders that go through a cart, on the other hand, are recorded in the same format from day one: name, product, amount and date, all in one place you can search.

That doesn't mean you have to rush. But if your chat count says it's time, starting a year earlier means a year more of customer history you get to keep.

What Kind of Cart a Chat-Based Shop Should Start With

A shop putting up price tags for the first time doesn't need to build a whole supermarket. For a shop that has always sold through chat, I think the first version needs to shrink piles one and four before anything else.

Product pages need to show the price, sizes and the colours actually in stock, so the basic questions disappear from your chats. The cart needs to accept the transfers your customers already use, and to store the name and address customers type in themselves, so the admin doesn't have to retype them. And every product page should keep a chat button, for customers who still want to talk before deciding.

A cart and a chat can live side by side. You don't have to pick one. For the details of which features you really need and which can wait, see my separate piece on which features a first e-commerce website actually needs, and which are wasted budget. If you'd like to see the scope and starting price of a website written from scratch for your shop, have a look at TumWebSME's website packages (prices on that page exclude 7% VAT).

A Real Option for Shops Selling Through LINE

If your shop already sells through a LINE Official Account, there's an option worth knowing about before you talk about a website. LINE offers a service called MyShop, which adds a storefront with a cart directly inside your shop's LINE OA. Its fees and promotions change from time to time, so I haven't put any figures in this article. If you're interested, check the latest terms with LINE for Business directly before you decide.

For a shop whose customers are almost all on LINE already, MyShop may well be the easiest first price tag to put up, because it sits where your customers already are. I'll say that plainly. The thing to weigh is that this storefront lives inside LINE's system and runs by LINE's rules. Shops that want people who aren't yet LINE friends to find their products through Google, or that want to design the whole ordering flow themselves, tend to start looking at their own website as the next step.

One real project our team built is Namtan Zhongli, a shop selling Thai products to Thai people living in Taiwan. Each product page shows the price in Taiwan dollars, the number of items left in stock, and an add-to-cart button right there. Pile-one questions like "How much is it?" and "Do you still have it?" are answered before a customer ever needs to message. For customers who still have questions, the shop gathers its contact channels (Facebook, LINE, TikTok and phone) in a contact section, and states clearly that it accepts both cash on delivery and transfers in Thai baht or Taiwan dollars. The case is also mentioned in our article on whether online stores really need a website. I don't have before-and-after sales figures from this shop that I can share, so I'm only describing what you can see on the live site, not claiming results I can't verify.

A Two-Week Plan Before You Decide

If you'd like to decide based on your own shop's numbers, try this.

1. Days 1-7: Sort your chats into the four piles every day, not just once. Your Monday and your Saturday may look very different.

2. Days 1-7: Ask your admin to note roughly how many hours a day go to pile one. There's no need to track it to the minute.

3. Day 8: Gather the pile-one questions and rank which ones come up most often. These are what your product pages need to answer.

4. Day 9: Multiply those hours by your admin's actual wage to get the monthly cost of pile one.

5. Days 10-14: Compare your options with all the numbers in hand: start with MyShop in your LINE OA, or talk to a web team about your shop's own website.

How to Measure Results Once the Cart Is Live

About a month after the cart goes live, count your chats the same way again. The first number to check is whether pile one has shrunk. If questions about price and shipping are still coming in at the same rate, your product pages aren't answering them yet, or customers don't know those pages exist.

The second number is the share of orders coming through the cart compared with chat. In the early weeks the cart may not get many orders, and that's normal. Your regular customers have been ordering by chat for a long time. Start attaching product page links to your chat replies, and customers will gradually see for themselves that they can order directly.

The third number is orders placed outside working hours. If orders start arriving at 10pm or early in the morning without anyone having to answer a chat, those are the customers who used to end up in pile three.

If two or three months pass and none of these three numbers has moved, don't conclude that a cart isn't right for your shop just yet. Open a product page on your own phone and place a real order. Often the problem is a small detail, like shipping costs that only appear at the very end, or a transfer button that's hard to find. Both match the top reasons shoppers abandon carts in the Baymard data mentioned above.

And don't forget to ask your admin how it feels. If they say they now have more time for customers with the harder questions, the cart is doing its job in the right place.

Closing Thoughts

The auntie at the corner shop doesn't need to go anywhere. On the day the shop puts up price tags, she's still sitting behind the counter, still asking after her customers' families. She just doesn't have to say how much the fish sauce costs a hundred times a day anymore.

A shop that sells through chat is the same. A cart doesn't replace your admin. It helps with the queue at the counter. If you count your chats and find pile one is still small, your shop may not need to do anything yet, and that's a perfectly good answer. But if you find your admin's day disappearing into the same questions, get in touch and tell us your four pile counts. We'll help you work out what kind of price tags your shop should start with.

Follow TumWebSME

Stay updated on web development and digital marketing:

• Facebook: TumWebSME รับทำเว็บไซต์ธุรกิจ

• Instagram: @tumwebsme

• TikTok: @tumwebsme

• YouTube: TumWebSME

Contact and Service Enquiries

• 088-983-9386 (Khun Ploy)

• 099-856-3198 (Khun Saennan)