What the WhatsApp 24-hour window does to your follow-up
On WhatsApp, a business can send a free-form message only inside a 24-hour customer service window that opens when the customer messages you. Once it closes, the only thing you can send is a pre-approved template. That single platform rule rewrites the follow-up habits a Sales QA review of social conversations normally checks for, because a follow-up policy that works everywhere else runs out of road at hour 24.
Most teams file that as a deadline. It behaves more like a budget, and it is usually spent on messages that give the buyer no reason to write back.
What is the WhatsApp 24-hour window?
It is the customer service window on the WhatsApp Business Platform: a 24-hour timer that starts when a WhatsApp user messages or calls your business, and restarts every time that user writes again. Four rules define it, and all four are set out in Meta's developer documentation.
- The timer starts when a WhatsApp user messages or calls the business.
- Every new message from that user resets the timer to a fresh 24 hours.
- While the window is open, the business can send any service message type with no pre-approval.
- Once the window closes, the business can only send pre-approved template messages.
Read that as a salesperson, not a developer. Your permission to write freely to a buyer is granted by the buyer, one message at a time, and it expires. Whatever your follow-up policy says, on WhatsApp the platform gets a vote.
Why is the window a budget rather than a deadline?
Because you do not control the refill and cannot pause the drain. The buyer funds the account, the balance runs down whether you use it or not, and the only way to get more time is to write something they want to answer. The quality of your message and the length of your access are the same problem.
What does each move inside the window actually buy you?
Only one common move buys anything back, and it is the one that ends with something the buyer can answer. Here is the same 24 hours spent five ways.
| What you send inside the window | What it costs | What you get back |
|---|---|---|
| An automated acknowledgment, then nothing until morning | Every hour it runs while nobody reads the thread | A notification the buyer has no reason to answer. No reply, no refill. |
| A price, a policy or a long explanation with no question attached | Everything after it | Something to read. Nothing to answer, so the thread stops on your turn. |
| A careful reply written at hour 23 | The last of the allowance | A message that lands, on a clock still counting from the buyer's last message. |
| A reply ending on something the buyer can answer in one word | A single message | An answer, and that answer resets the timer to a fresh 24 hours. |
| Nothing at all until hour 25 | All of it | A thread you can now reopen only with a pre-approved template. |
The fourth row is the only one that pays for itself, and what belongs in it is the subject of turning a DM into a booked appointment. Here it stops being a matter of style, because the buyer's answer is the only thing that puts hours back on the clock.
Does a business reply reset the WhatsApp 24-hour window?
No. Only a new message from the customer does, so three careful replies in a row extend nothing: the window still ends 24 hours after the last time the customer wrote.
That makes the question at the end of your reply the only renewable resource in a WhatsApp thread. Let me know what works for you is weak selling on any channel. Here it is also a decision not to refill the clock.
The corollary is the useful half. A thread whose last message is the customer's holds a full, untouched window, and those are exactly the threads that look finished from the inbox. Nobody chases a conversation where the other person spoke last.
The example Sales QA report on the Reploom homepage has one thread that survived on that rule by accident. Nolan Price wrote to a studio in the evening, got an automated line saying it was closed until 7 AM, then messaged again on his way home. That second message handed the studio a fresh 24 hours before anyone there was awake. He reset his own window, and no process at the studio did it for him. Sample data from one account, not a benchmark.
What does spending the window badly look like in one thread?
It looks like a thread where every message is accurate. In the same example report, Layla Perez is a WhatsApp lead scored 81 with pricing intent, flagged under the pattern Answered, but no next step.
She wrote three times, so the studio was handed three full windows. It spent the first on a price, the second on a scheduling fact, and the third not at all. Her last message, I am still interested, but I am not sure how to reserve the consultation, is where the thread stops. Two accurate replies, no question in either, and the window that was still wide open is the one nobody used. A day later, the line that would have closed her, a named time somebody offered to hold, is no longer something anyone can send.
How fast does a quiet WhatsApp buyer have to be noticed?
Within the same day, because the platform clock is shorter than most teams' review cadence. The 24 to 48 hour follow-up rule for a buyer who goes quiet after a price was written as a sales habit rather than against a platform rule, and on WhatsApp only the first half of that range is still free-form. A weekly rhythm fails the same way, larger: a Friday conversation read on Monday is read after its window closed.
Evening volume is where that bites hardest. In the example report, 31% of leads arrived after 20:00, and first replies averaged 41m inside working hours against 1h 23m outside them. Sample figures from one account, but the shape is the point: a buyer who writes at 21:30 owns a window that expires at 21:30 tomorrow, and most of the gap is spent asleep. Setting a target per time window is the subject of reply speed as a growth lever.
How do you know which WhatsApp threads are about to age out?
Sort them by the timestamp of the customer's last message, not by the latest activity. Nearly every inbox sorts by the latter, which is usually your own reply, so the threads closest to closing sit near the top looking handled.
- Read by the customer's last inbound message. An inbox sorted by recent activity is sorted by the wrong column.
- Start where the buyer spoke last. Those threads carry the most open access and the clearest unanswered question.
- Pull anything past twelve hours whose last exchange holds no time, no question and no offer. Half the allowance is gone and nothing invites a reply.
- Clear the list before you stop for the day, one owner each. Anything carried overnight is carried across the deadline, not to tomorrow.
Friday is its own case. A message that arrives at 21:00 on Friday runs out at 21:00 on Saturday. If nobody works the weekend, the free-form reply to that buyer either goes out before the weekend or it does not go out.
What Reploom does here, and what it does not
Reploom does not manage the window. It does not send template messages, does not submit templates for approval, does not run a countdown beside your threads, and will not warn you before one closes. It does not draft, send or approve replies either, it is not an inbox and not a chatbot, and it does not replace the people answering messages. The rule is Meta's and the work is yours.
What it does is review what already happened. Reploom is AI Sales QA for social conversations: it reads Instagram, Facebook and WhatsApp messages, comments and mentions after they happen, finds the leads a team lost, and groups repeated gaps into named patterns in a weekly report with the source conversations as evidence. On WhatsApp, monitoring starts when the number is connected, so how far back a first review reaches depends on the history Meta makes available.
Unanswered leads and overdue follow-ups also surface in a live queue measured against the reply target you set, a sales clock you chose rather than the platform's. The report itself names the pattern underneath: how many buyers were answered with no next step, how many stalled after a price, and which conversations those were.
The example report's headline is 7 sales opportunities stalled this week. It counts stalls, not windows. To see which of your own conversations went quiet on the customer's turn, Book a Demo.