WhatsApp Business API Pricing: What a 2027 Broadcast Programme Really Costs a Singapore SME

A WhatsApp broadcast programme for a Singapore SME with a few thousand opted-in contacts typically lands between S$700 and S$1,200 a month — and roughly nine-tenths of that is marketing template messages, not platform fees. Since Meta moved to per-message pricing in mid-2025, you are charged for every template message you send, priced by category (marketing, utility, authentication) and by the recipient's country. Service messages — your replies inside an open customer conversation — are free. That single change reshapes the economics: the lever on your bill is no longer how many conversations you open, it is how many names you broadcast to and how often. So before you sign anything for 2027, you need three artefacts of your own: a message-category map, a deduplicated opt-in list, and a cost-per-outcome figure. A vendor quote is not a budget.

How does WhatsApp Business API pricing actually work?

Your real cost sits in three layers, and most SMEs only ever see the middle one.

Layer one is Meta's per-message charge. Every template message you initiate is billed at a rate set by category and recipient country. Marketing is by far the most expensive category — commonly five to ten times the cost of a utility or authentication message in the Singapore rate card. This layer is non-negotiable and scales linearly with your list size.

Layer two is your Business Solution Provider. Whoever gives you the inbox, the template editor and the campaign scheduler charges either a monthly platform fee, a per-message markup on top of Meta's rate, or both. Platform fees for small teams generally run S$80 to S$250 a month. Markups are where quotes diverge most, and where you should ask for the number in writing.

Layer three is build and maintenance. Connecting WhatsApp to your CRM or POS, writing and submitting templates for approval, and keeping the opt-in list clean is real work. Budget it as a recurring line, not a one-time project, because template rejections and catalogue changes will keep arriving.

Which messages are free and which ones do you pay for?

Getting this map right is the single biggest cost decision you will make, because the same business event can be routed as a free message or a paid one depending on timing and category.

• Service messages — free. Anything you send inside the 24-hour window after a customer messages you. Answering enquiries, confirming a booking a customer just asked about, handling a complaint thread: no per-message charge.
• Utility templates — paid, but free inside an open window. Order confirmations, delivery updates, appointment reminders, invoice notices. If the customer has messaged you in the last 24 hours, these ride the open service window at no cost. Outside it, you pay the utility rate.
• Authentication templates — paid, low rate. One-time passwords and login codes.
• Marketing templates — paid, highest rate. Promotions, re-engagement, new arrivals, anything persuasive. This is your bill.
• Click-to-WhatsApp ad entry points — free window. When a customer starts the conversation from an ad or a Facebook/Instagram entry point, you get a free window to respond. You still pay for the ad.

If your order confirmations are being sent as marketing templates because nobody categorised them properly, you are paying several times over for messages that could have been free.

What does a realistic monthly programme cost?

Take a retail or services SME with 4,000 opted-in contacts. The rates below are illustrative round numbers for modelling only — pull Meta's current Singapore rate card and your BSP's markup before you commit a budget line.

Call it S$10,600 a year. Now put it against an outcome: if those 8,000 marketing messages convert at 1.5% on a S$60 average order, that is 120 orders and S$7,200 of revenue a month against S$885 of messaging cost. The programme works. But notice where the sensitivity lies — cutting your platform fee in half saves S$75, while cutting one monthly broadcast saves S$360. Frequency and list quality are your cost controls, not vendor negotiation.

Why is the first quote always lower than the real bill?

Four reasons, in order of how often they bite.

Duplicate and dead numbers. A list assembled from POS exports, a booking system and two spreadsheets will carry 10 to 20% duplicates and numbers that no longer have WhatsApp. You pay to attempt delivery to contacts who will never read anything. Normalise every number to E.164 format (+65...) and deduplicate before you count your list.

Category drift. Templates get written in a hurry, submitted as marketing because that approves fastest, and never revisited. Audit your template library quarterly and reclassify anything genuinely transactional.

Minimum commitments. Some BSP contracts bundle a message allowance you will not use in month one. Match the tier to your actual send volume, not your ambition.

Testing in production. Every rehearsal send to a staff list is billed. Keep a small, named test group and know what it costs.

Does PDPA consent change what you can broadcast?

Yes, and this is where the cheap-list shortcut becomes an expense. Singapore's Do Not Call provisions are aimed at voice calls, SMS and fax to local numbers, so WhatsApp broadcasts sit outside the DNC Registry. That is not a free pass. The PDPA's consent obligation still applies to the personal data you are using, which means you need a defensible record of how each contact opted in, and you must honour withdrawal promptly and permanently.

Practically, that means your contact record needs three fields it probably does not have: opt-in source, opt-in timestamp, and opt-out timestamp. If a complaint arrives in 2027, those three fields are your answer. Scraping a WhatsApp group's member list for a broadcast is the exposure, not the shortcut.

What should you fix in your system before scaling?

1. Normalise all phone numbers to E.164 and deduplicate across every source system.
2. Add opt-in source, opt-in timestamp and opt-out timestamp to your customer master record.
3. Category-tag every template and route transactional events as utility, timed to ride open service windows where possible.
4. Set a hard per-campaign spend cap in the BSP console so a mis-targeted send cannot run away.
5. Instrument the funnel beyond delivery: delivered, read, clicked, ordered. Delivery rate is not a result.
6. Keep the CRM as the single source of truth. A list that lives in one person's phone contacts is not an asset you own.
7. Name an owner for the template library, and a second person who can run a broadcast. If only the boss can send, you have replaced a marketing channel with a person-shaped bottleneck.

Do those seven things and the pricing question answers itself. Skip them and you will spend a year paying marketing rates to broadcast duplicate records with no consent trail and no attribution — which is expensive at any rate card.

Frequently asked questions

Is WhatsApp Business API cheaper than SMS for a Singapore SME?

For transactional messaging, usually yes — utility and authentication templates typically undercut local SMS rates, and service replies are free where SMS never is. For bulk promotional sending the gap narrows, because marketing templates carry the highest WhatsApp rate. The real advantage is not unit price but read rate and the ability to hold a two-way conversation in the same thread.

Do we need the API at all, or is the free WhatsApp Business app enough?

If you are under a few hundred contacts, one person handles replies, and you are not automating anything, the free app is enough and costs nothing. You move to the API when you need multiple staff in one inbox, automated order or appointment messages triggered from another system, or broadcasts beyond the app's narrow list limits. The trigger is operational, not revenue size.

How far ahead should we budget for messaging in 2027?

Model twelve months at your intended send frequency, then add 30% for list growth and re-sends. Review the figure quarterly against cost-per-order rather than cost-per-message — a programme that looks expensive per message can still be your cheapest acquisition channel, and a cheap one sent to a dirty list is pure waste.