Alimtalk

For a long time the only ways to deliver a message that had to arrive were text and email. A failed payment or a confirmed reservation is something a customer needs to know within hours, and neither channel kept up. Text messages ran out of characters as soon as a link went in, and a short message from an unknown number with a link attached is the kind of thing people decline to tap. Email had room to explain but sank down the inbox.

App push was the third road, with a condition attached: it reaches people who installed the app and left notifications on. Customers who only use the web, and customers who muted the app months ago, sit on the list unreachable.

Pagedeck hit this directly. A team plan payment failed against a card limit, and the notice went out by email. The admin opened it three days later, by which point the workspace had dropped to read-only. Ravi checked and found the email had been delivered without trouble — it simply was not read. That is the moment the gap between sent and received becomes a number someone has to answer for. Alimtalk is the answer that routes those notices through the messenger nearly everyone in the country opens several times a day.

It solves reach first. It doesn't matter whether the recipient installed an app or added the sender's channel as a friend. A phone number collected at sign-up is enough for the message to land in that number's KakaoTalk, sitting in the conversation list alongside messages from actual people rather than queued behind newsletters.

In return, it accepts a hard limit on content: informational messages only. Order confirmations, shipping status, failed payments, booking confirmations, verification codes — things a customer needs to know because of a transaction they entered into. Promotional lines, discounts, and new-feature announcements are not allowed.

Sending one takes five steps.

1. Open a KakaoTalk channel. This is the sending account, registered under the business entity.
2. Sign with a messaging intermediary. Messages go through that provider rather than straight to Kakao.
3. Register a template and submit it for review. The whole sentence is written out in advance, with only per-recipient values such as a name or an amount left as variables.
4. Once approved, fill the variables and send. Not a character of the approved text can change.
5. Configure a fallback so that anything undelivered goes out as a text message instead.

Step three defines the character of the channel. Review usually takes about two business days, so a line thought up this morning cannot go out this afternoon. That makes it a natural fit for trigger messages, which fire automatically the moment a defined event occurs — a payment failure happens without warning, but the sentence sent in response can be written long before.

Cost runs around $0.006 per message, below a text message. Against a long text with an image attached, the gap widens to roughly fivefold. For a service sending tens of thousands of notices a month, that difference alone can decide the channel.

One consequence catches teams off guard. Because the approved sentence is frozen, anything that changes often — a price, a policy line, a feature name — either lives in a variable or forces a new round of review. Teams that plan for this keep a small set of durable templates instead of writing a fresh one per campaign.

Four other channels carry the same kind of notice.

1. Text messages — no review, so wording decided today goes out today. Two to three times the cost per message, and tight on length. Its main role here is catching what Alimtalk fails to deliver.
2. Email marketing — room to write and effectively no per-message cost. Too slow for anything the customer needs to act on today.
3. App push — free to send and visible on a locked screen. Reaches only people who have the app, and once someone turns notifications off there is no way back.
4. In-app messages — shown to people already inside the product. Useless for bringing back someone who left.

There is a way to send promotional content through KakaoTalk, through a separate product aimed at people who added the channel as a friend. The audience shrinks to those who opted in, and any of them can drop the channel at any time.

"This copy is promotional, so it can't go out as Alimtalk."

A line built around a discount will be rejected in review.

"Template review takes two days, so the wording has to be final today."

Deciding today for a send tomorrow is already too late.

"I'll send Alimtalk first and set failures to fall back to text."

Only the numbers that can't receive it on KakaoTalk will cost us a text message.

From the week Pagedeck moved its trial expiry notice. Ravi decided to switch the three-days-left reminder from email to Alimtalk, and wrote a line offering 20 percent off for paying now. The template came back rejected, because the discount counted as promotional.

He split it in two. A template carrying only the expiry date and the days remaining went back in and was approved, while the discount went out by email separately. The expiry notice got read far more often than the old email version, and the open rate on the discount email stayed about where it had been. Changing the channel improved whether the notice arrived in time. It did not improve persuasion.

The most common mistake is treating it as a cheap marketing channel with great engagement. Looking only at read rates makes that easy to believe. But the only thing this channel carries is what a customer needs to know about their own transaction. Promotional copy dressed up as a notice usually fails review, and when it slips through, accumulated reports can restrict the sender's ability to use the channel at all. People read these messages precisely because they aren't ads, so putting an ad there removes the reason they got read.

The second is assuming that informational means consent is unnecessary. It's true that these messages go out without prior marketing consent, but the basis is the transaction. The customer ordered or paid, the message concerns that order, and the phone number was collected for it. Send to a number with no such relationship, or send content unrelated to the transaction, and the message crosses into territory that requires opt-in consent. What comes back across that line is not weak performance but a fine.