"We don't have a system. We just have a LINE group. Everything gets said in there."
I've heard this too many times. Usually it's the owner saying it in the meeting room, laughing, half embarrassed and half defiant. On the defiant part, he's right. I'm Johnny Yang, from Brightstream Technology in Taiwan. I've led implementation work for seven years: 252+ projects across 32+ industries. In those seven years I have never once convinced a client that "LINE is wrong," because LINE was never wrong. When a company has no processes at all, it's the only thing that lets everyone see the same thing at the same time.
The problem comes later. When the company grows from eight people to twenty-five, from one production line to three plants, from the owner remembering everything to the owner no longer being able to, the LINE group goes from helping you to lying to you. Everything looks like it's working. Messages keep flowing. Everyone responds. But the things that actually need to be remembered are all sinking.
So here's the conclusion up front: don't try to drive LINE out of your company. That's a war you won't win, and you don't need to fight it. What you need to do is draw a line. Anything that's "said and done, nobody needs to look it up later" stays in LINE. For that, it's the most efficient tool in the world. Anything that someone will later need to look up, bill for, assign blame for, or hand over must never exist only in LINE. The test is one sentence: three months from now, will someone ask "who said that, and when?" If the answer is "possibly," it has to leave LINE and land somewhere with fields, an owner, and a status. You don't need to buy anything to start drawing this line. This afternoon you can split everything your company does into two piles on a whiteboard.
Below I break down the three risks one by one, each with what it actually looks like on the floor. Talking about risk in the abstract is useless. The owner nods, goes back, and changes nothing.
First, admit why LINE actually works
In a lot of my projects, the owner is defensive from the start, assuming the consultant is there to scold them for doing things the crude way. I usually make it clear first: a LINE group holding a company together for years isn't luck. It solves the two hardest problems in implementation.
First, the learning cost is zero. When any system goes live, the biggest resistance is never missing features. It's that people on the floor don't want to learn it. Ask a master technician with fifteen years on the job to remember a username and password and click through four layers of menus to find the work-reporting screen, and he'll resist in his own way: he'll keep doing the work and just not fill anything in. LINE needs no training. A sixty-year-old mold maker can use stickers. A lot of consultants underestimate how valuable that is.
Second, it's already there. The most expensive hidden cost of implementation is asking everyone to communicate somewhere new. If even one person doesn't switch, information splits into two channels, and people naturally drift back to the one with more people in it. LINE wins because it never had this problem. Employees' personal lives are already on it.
So when someone tells you "running a company on LINE is unprofessional," that statement isn't worth much. The real question isn't whether it's professional. It's whether it can still carry the company you are now. LINE can carry communication. It can't carry records. When a company is small, those two look almost identical. As it grows, they split.
Risk one: messages sink, and what sinks isn't the message, it's the responsibility
This is the most talked-about risk and also the most underestimated. People think the problem is "too many messages, so things get missed." The real problem is that LINE sorts by time, and a company runs by status.
I saw one scene I still remember clearly. A salesperson posted in the group that the customer wanted a spec change. The warehouse keeper replied with a sticker. Production control said "I'll take a look." The next forty messages were about which restaurant to book for the year-end banquet. Three days later the goods shipped to the old spec. When we traced it back, everyone was innocent: sales said it, production control replied, the warehouse saw it. Every message was there. Nobody owned the outcome.
The problem is that "I'll take a look" isn't a status. It's a sentence. A status in a system has a life cycle: pending, in progress, done. If it isn't finished, it stays lit up and keeps poking you in the eye. A sentence in LINE has no life cycle. The moment it's sent it starts drifting up, and by the next morning nobody remembers it ever existed.
Worse, scrolling up to find a message can't be audited. You can scroll for ten minutes and say "I can't find it," and nobody can prove whether you really looked. So "I didn't see it" becomes an excuse that always holds, and any excuse that always holds will get used. That's not a character problem. It's a design problem.
There's a simple test for whether you've hit this risk: ask someone right now to reconstruct the full handling of a rush order from last month. If their first move is to open LINE and scroll up, your company is already running on memory. This is really the same problem as still using Excel to run the company, in a different form. One has data scattered across files. The other has it scattered across conversations.
Risk two: the group can never tell you who owns this
The second risk is even less visible than the first, because normally nothing goes wrong. When something does, it's big.
A LINE group is, by nature, a broadcast. Say something in the group and you're saying it to everyone, which means you're saying it to no one in particular. Psychology has a very intuitive explanation for this: the more people there are, the less responsible each person feels. Post "can someone handle this" in a twenty-person group, and the most likely result is that all twenty assume someone else will.
The fix I've actually seen is the owner starting to put names in front of messages: "@Ming, you handle this." It does work. But it makes assigning responsibility depend entirely on one person's attention. If the owner remembers to tag, the work gets done. If the owner is out visiting customers that day and forgets, nobody does it. You've built the company's dispatch system on top of one person's head. In a ten-person company that's completely workable. In a thirty-person company it's the setup for a disaster.
There's another, worse layer: LINE has no "decline" action. When a system assigns a task, the assignee can send it back, mark it as unable to handle, write a reason, and all of that leaves a trace. In LINE, you tag him, he doesn't reply, and you don't know whether he didn't see it, saw it and didn't want it, or thought you meant someone else. The handoff point for responsibility is blurry, and every management problem happens at the handoff point.
This is especially deadly for quotations and collections. Who promised the customer that discount, who agreed to push back the payment date on their post-dated check, who said we could ship first and do the paperwork later. When these conversations live in LINE, reconciling accounts afterward turns into each side telling its own story. When a bill can't be settled, the root cause is usually not a miscalculation. It's that the original commitment was never recorded anywhere both sides accept.
Risk three: when someone leaves, the company's memory walks out with their phone
This is the one the fewest people think of beforehand, and the one that hurts most afterward.
Who holds the data in a LINE group? Not the company. Every individual phone. The day an employee leaves, they exit the group or you remove them, and for the company those conversations effectively no longer exist. More often there's no group at all, just one-on-one chats between a salesperson and customers. That salesperson holds the history with thirty customers, the quotation trail, the buyer's personal preferences, which customer you can't call on Mondays, all in their personal LINE. They leave, and it all goes to zero in a second.
I saw a trading company where, after a senior salesperson left, the new hire spent nearly six months getting to know the customers all over again, and lost several orders that would have renewed without question. The owner asked me something then that stuck with me: "All these years, was I building a company, or was I just paying him?"
That question was dead on. The company's asset should be the customer relationships. But if every interaction lives only in a personal messaging app, what the company actually owns is its employment relationship with that salesperson. When the employment ends, so does the asset. This is why so many owners only discover at succession time, or when bringing in new people, that there's nothing to hand over. All the know-how is word of mouth plus chat logs on personal phones.
Incidentally, there's a compliance side to this too. Personal data protection, trade secrets, evidence in labor disputes: all of them assume you control your records. You don't control what's on an employee's personal phone. Normally nobody asks. The day someone does, you'll have a hard time answering.
So how do you actually draw the line
That's the three risks. Here's the point: the answer is not "ban LINE." I have never seen a company successfully ban LINE. It usually doesn't last two weeks, because people on the floor need speed, and LINE is the fastest. Ban it and they'll keep talking in private groups. You just won't see it anymore.
The real answer is division of labor. LINE handles speed. The system handles accuracy. Those two jobs were never meant for the same tool.
| Situation | Stays in LINE | Must leave LINE |
|---|---|---|
| Quick notices (power outage this afternoon, owner running late) | Fine | — |
| Photos from the floor reporting progress, confirming it's the right box | Fine (but the conclusion goes into the system) | — |
| Orders, spec changes, delivery commitments | Not fine | Needs an order number and version |
| Leave, overtime, shift cover | Not fine | Needs an approval record |
| Quotations, discounts, payment terms | Not fine | Needs an approver |
| Customer history, contact preferences | Not fine | Must be a company asset |
| Quality issues, customer complaints | Can be reported here first | Must be ticketed and tracked to closure |
Don't copy this table as is. Understand the test behind it: if someone might come back and ask about it three months from now, it can't live only in a conversation.
In practice, the most effective approach is to treat LINE as the entry point, not the destination. People on the floor keep talking, taking photos, and reporting in LINE. Don't touch that habit. What changes is the second half: someone is responsible for moving the conclusions reached in LINE into a place with fields. At first that person might be the sales support staff, and the tool might be a shared form. That's completely fine. The point isn't how sophisticated the tool is. It's making "conclusions must leave LINE" a rule.
When the manual copying gets heavy enough that someone starts complaining, that's when to consider connecting the entry point to a system. Not earlier. Too many companies buy tools before the rule exists, nobody uses what they bought, and everyone keeps running things in LINE. Same logic as which process to fix first: get the order wrong and the money is wasted.
Three things you can do tomorrow
First, open a new group whose name says it does exactly one thing, for example only "today's shipment confirmations." A single-purpose group doesn't sink, because there's no noise in it. Keep the old everything-goes group for chatting and quick notices. Leave it alone.
Second, pick the most painful step, usually order changes or leave requests, and set a rule: starting next Monday, after this is discussed in LINE, someone must enter it in a form and reply "logged" in the group. Do just this one step. Don't change everything at once. Change everything at once and everything reverts.
Third, ask your sales manager: if someone quit right now, how much on their phone could the company never get back? The answer doesn't need to be precise. Just watch how uncomfortable he looks. That level of discomfort is your current risk exposure.
LINE isn't the problem. Using LINE as the company's memory is. If you want to know which of your processes should be moved out of the chat window first, book a free 30-minute diagnostic with me. I'm not selling anything. I'll just help you draw the line.




