Skip to main content
RoutMe
All posts
WhatsAppUpdated 20 October 20263 min read

WhatsApp per-message pricing, explained without the spin

Meta bills WhatsApp per message, not per conversation. Here is what that changed, what became chargeable in October 2026, and how to estimate a bill before you send anything.

SM

Sushma Madan

WhatsApp platform

Since 1 July 2025, WhatsApp Business bills per message rather than per 24-hour conversation window. Most teams still budget with the old mental model, and the two produce very different numbers for the same traffic. This post is the version we wish existed when we were doing the migration maths ourselves.

What actually changed

Under conversation-based pricing you were charged once when a 24-hour window opened, and everything inside that window was free. You optimised for opening as few windows as possible. Under per-message pricing, each template message you send is billed on its own, and the category of that template decides the rate.

The practical consequence is that chatty automation got more expensive and tight automation got cheaper. A flow that sent five templates inside one window used to cost the same as a flow that sent one. Now it costs roughly five times as much.

The categories you are billed on

  • Marketing: promotions, offers, re-engagement, anything designed to bring someone back. The most expensive category in every market, and the one Meta rate-limits hardest on quality.
  • Utility: transactional follow-ups tied to something the customer already did: order confirmations, shipping updates, payment receipts, appointment reminders.
  • Authentication: one-time passcodes and verification. Priced separately, usually the cheapest template category.
  • Service: your replies inside a customer-initiated conversation, sent as free-form messages rather than templates.

How to estimate a bill honestly

Rates vary by market, and the market is determined by the customer's phone number, not where your company is registered. An India-heavy list and a Brazil-heavy list with identical volume produce very different invoices. So estimate per market, not in aggregate.

  1. 1Split your contactable audience by country code. Take the top five markets; the tail rarely moves the total.
  2. 2For each market, count template sends per contact per month, separated into marketing, utility, and authentication.
  3. 3Multiply by that market's rate for each category and sum. This is your floor.
  4. 4Add your expected service-message volume, inbound conversations multiplied by the number of turns your agent takes to resolve one, then subtract the free tier of 1,000 delivered service messages per business phone number each month. Everything above that is billed.
  5. 5Apply a quality discount of nothing. Do not assume you will land in a cheaper tier because your templates are good.

The number that surprises people is step four. Teams model their template spend carefully and then discover their support agent averages nine turns to resolve a question that a well-scoped agent resolves in three.

Where the savings actually are

Per-message pricing rewards precision, and precision is mostly an engineering problem rather than a negotiation one.

  • Resolve in fewer turns. An agent that can read the order status directly answers in one message. An agent that cannot has to ask for the order ID, wait, then look it up. That is three billable messages instead of one now that service messages are billed.
  • Stop sending utility templates nobody reads. Shipping updates at every hop felt free under conversation pricing. They are not free now.
  • Move re-engagement out of marketing where it legitimately belongs elsewhere. A genuine order follow-up is a utility template, and miscategorising it upward is a common and expensive habit.
  • Cut the confirmation ping-pong. 'Did that answer your question?' is a billable message that rarely changes an outcome.

What we do not know

The current rate card takes effect 1 October 2026, and Meta revises it periodically. Anything you read about specific per-message rates, here or anywhere else, should be checked against Meta's published pricing before you put it in a budget. We link the source rather than reprinting numbers that go stale.

The structural points above outlast the rate card. Bill per message, price per market, and count your turns.

  • Pricing
  • WhatsApp Cloud API
  • Planning
SM

Sushma Madan

WhatsApp platform

Writes about templates, pricing, and getting things approved on the WhatsApp Business Platform.

See it on your own data

Connect a system, drop an agent node into a flow, and watch it answer from what is actually true in your systems right now.