The WhatsApp 24-hour customer service window explained

Every time a customer sends a message to a business on WhatsApp, it opens a customer service window that runs for 24 hours from that message. Inside that window, the business can send free-form, plain-text replies — no template, no pre-approval, no category restrictions — exactly like a normal conversation. This is the mechanism that makes WhatsApp usable as a genuine two-way support channel rather than a one-directional broadcast tool: a customer asking a b2b sales desk about pricing gets a normal, unstructured reply, not a stiff pre-written template forced into every response.

The window resets every time the customer sends a new message. If a lead writes again six hours into an open window, the clock restarts from that new message, giving the business another full 24 hours of free-form reply capability. This is why an active, ongoing conversation effectively never closes the window as long as the customer keeps replying — the closing only becomes relevant the moment the customer goes quiet and the business still has something to say.

What closes the window is simply time passing with no new inbound message from that customer. There is no warning bell inside WhatsApp itself that tells an agent 'this window closes in ten minutes' — the window closes exactly at the 24-hour mark from the customer's last message, silently, and any reply attempted after that point is treated by the API as an entirely different kind of send, one that requires an approved template rather than plain text.

This is the detail that catches businesses out in practice: a free-form message sent after the customer service window has closed does not bounce back with an obvious, readable error explaining what went wrong. It fails at the API level, and depending on how the business's inbox software surfaces that failure, an agent may see nothing more informative than a generic send failure, or may not notice a failure at all. The customer, meanwhile, receives nothing — no reply arrives, and they have no way of knowing whether the business ignored them or the message simply never went through.

A b2b sales rep replying to a lead's pricing question four days after the lead's last message is the textbook version of this failure: the free 24-hour window closed on day two, and the plain-text reply typed on day four never reaches the lead at all. From the rep's side, it can look like the message sent successfully — the local inbox shows it as sent — while the actual delivery to WhatsApp's servers was rejected because no valid template was used for a send outside an active window. This gap between 'looks sent' and 'was delivered' is why relying on visual confirmation alone, without an explicit window-status indicator, is a genuine risk at any real reply volume.

The fix once a window has closed is not complicated in principle — send an approved template message instead of plain text, and that template send, once delivered, does not itself reopen the free-form window; it is simply a separate paid message that reaches the customer. Only a new inbound message from the customer reopens the window for free-form replies again.

Separate from the customer-initiated 24-hour window, a different 72-hour window opens when a customer arrives through certain qualifying entry points — most notably, clicking a Click-to-WhatsApp advertisement on Facebook or Instagram, or tapping a call-to-action button linked to WhatsApp on a Facebook Page. For that specific entry, WhatsApp treats messages within the following three days differently, allowing a business to message the customer without the same restrictions that would otherwise apply outside an active conversation.

The critical point, and the one most likely to cause confusion at any real volume, is that these two windows run on entirely independent clocks and do not move together. A customer who clicks a real estate agency's Click-to-WhatsApp ad opens a 72-hour free-entry-point window immediately, regardless of whether they've sent an actual message yet. If that same customer then writes 'tell me more about the 2BHK' a few hours later, a separate 24-hour customer service window also opens, starting its own countdown from that message. The 72-hour window can still be open while the 24-hour window has already closed, or the reverse can be true, with the ad-click window long expired while a fresh customer-service window is active from a more recent reply.

Treating these as if they were one window, or assuming whichever one is more generous automatically governs a given send, is the single most common source of confusion once a business runs any paid Click-to-WhatsApp advertising alongside its normal reply handling. The two need to be checked independently for any given contact at any given moment.

Understanding the window changes practical behaviour, not just billing awareness. A support team that knows a customer's window is closing in the next hour has a real incentive to resolve or at least substantively respond to that conversation before it lapses, rather than letting a half-answered query sit overnight and quietly become a template-only conversation by morning. A salon's front desk fielding a client's question about availability late in the evening, aware the window opened with that message and will close by roughly the same time the next day, has a concrete reason to close the loop that day rather than assuming 'we'll get to it tomorrow' costs nothing.

This also reframes what a reminder or follow-up message actually is, mechanically. A reminder sent well after a customer's window has closed is not a 'free follow-up' in the way it might feel casually — it is a template send, subject to template approval, category rules, and whatever billing applies to that category. A business planning any delayed follow-up sequence needs to build it around approved templates from the start, because by the time the delay has elapsed, the free window is almost certainly gone.

For a business handling any real conversation volume, the practically useful habit is having a visible, per-contact indicator of whether the customer service window is currently open, rather than relying on staff to mentally track when each customer last wrote in. A platform like Wavy that surfaces this status directly inside the inbox — flagging a reply as needing a template before an agent attempts to send a plain message that would otherwise fail silently — closes exactly the gap that catches businesses relying on memory alone.

One recurring misunderstanding is that sending a template message somehow reopens or extends the free-form window — it does not. A template delivered to a customer outside an open window is simply a paid message that reaches them; it does not grant the business a fresh 24 hours of free replies unless the customer themselves responds to it, at which point their reply is what reopens the window, not the business's template send.

A second misconception is that the window is per-business rather than per-contact — it is not. Each individual customer has their own independent window, tracked against their own most recent message, which means a business with an active support queue can have some customers inside an open window and others outside it simultaneously, with no single account-wide status that applies to everyone at once. A shared team inbox handling dozens of open conversations needs per-conversation window tracking, not one dashboard-level flag, because averaging or generalising across contacts here produces meaningless information.

A third misconception, understandable given how the terminology sounds, is that the window has anything to do with message categories or billing categories directly — it does not determine whether a message is marketing, utility, or authentication. The window governs whether a free-form, non-template message can be sent at all; the category system governs how a template message is billed once one is actually required. These are two separate mechanisms that happen to interact — a business needing a template because the window is closed still has to pick the correct category for that template, and getting that category wrong is an entirely separate problem from the window closing in the first place.

Once a business accepts that the window is a hard mechanical boundary rather than a soft guideline, the practical response is to design conversation-handling around it rather than treating every missed window as an unpredictable failure. An edtech company's admissions team, for instance, can build a habit of checking, at the end of each day, which open conversations are approaching the 24-hour mark with an unresolved query, and either resolving them or preparing an approved follow-up template for the ones that will close overnight.

For sequences that are inherently delayed by design — an appointment reminder sent the evening before a visit, a reconnect message sent days after a stalled enquiry — the right assumption is that the free window will not still be open by the time the message needs to go out, so building these as approved templates from day one avoids discovering the gap only when the first scheduled send silently fails. Treating window status as something to check per contact before any delayed message fires, rather than assuming a day or two's delay is still 'probably fine', removes the guesswork.

A b2b sales desk following up a stalled deal, or a real estate agent circling back to a lead who went quiet after a site visit, benefits from the same discipline: check whether the window is still open before typing a plain follow-up, and if it has closed, reach for an approved template built for that reconnect purpose.