How to Choose the Official WhatsApp API Provider Without Paying Twice

How to truly compare costs

Proposals are rarely comparable at first glance because each provider charges in a different place. To compare, always add up the same four layers:

• Tool monthly fee, per user, per number, or fixed.
• Meta message cost, which is the same for everyone and depends on the template category and recipient’s country.
• Margin per message: some providers pass on Meta’s pricing, others charge a higher per-message fee or sell credit packages.
• Exit and implementation costs: activation fees, minimum contract terms, fees to migrate the number or export history.

The margin per message is the most overlooked layer and the one that grows most with the operation. In a small campaign, it goes unnoticed; in an operation sending tens of thousands of reminders per month, it becomes the largest part of the bill. Categories and billing logic are explained in Meta conversation pricing.

The account comparison of two proposals

A network of three clinics sends about 9,000 utility messages per month outside the service window — confirmations and reminders — and 3,000 marketing messages. They received two proposals:

• Provider A: monthly fee of R$ 390, messages charged with a 40% margin over Meta’s pricing, 12-month contract.
• Provider B: monthly fee of R$ 790, messages at Meta’s pricing, no contract.

Assuming, for simplicity, an average Meta cost of R$ 0.05 per utility message and R$ 0.35 per marketing message, messages cost R$ 450 + R$ 1,050 = R$ 1,500 per month at list price.

• With A, messages cost R$ 2,100. Total: R$ 2,490 per month.
• With B, messages cost R$ 1,500. Total: R$ 2,290 per month.

The proposal with the lower monthly fee is R$ 200 more expensive per month — R$ 2,400 per year — and locks the network in for 12 months. If volume doubles with a fourth clinic opening, the difference exceeds R$ 800 per month. The per-message values here are illustrative; what matters is the method: always compare the total at the real volume, not just the monthly fee.

The 12-Point Checklist

Take these questions to every meeting with a provider. A vague answer to any of them is information:

1. Is the portfolio and account in my company’s name?
2. Is the company a recognized Meta solution partner or technology provider, or a reseller?
3. How much do I pay per message, and is it exactly Meta’s pricing or is there a margin?
4. Are message charges billed by Meta on my card, or by the provider?
5. Is there a contract term, penalty, or fee to migrate the number to another provider?
6. Can I export the conversation history if I leave?
7. Does registration use Meta’s official flow or ask for my Facebook password?
8. Can I connect a number already on WhatsApp Business without removing it from the phone (coexistence)?
9. Does the tool show the number’s quality and sending limit, or do I only find out when it drops?
10. How does support work when a template is rejected or the number is restricted? How quickly does someone respond?
11. Can multiple people handle the same number, with conversation responsibility and client history?
12. What happens to data if I cancel, and how quickly is it deleted?

Warning signs in proposals

• "We handle everything; you don’t need a Meta account." Likely translation: the account stays with them.
• Request for someone’s Facebook password from the company. The official flow never requires this.
• Promise of sending without an approved template or "no limits." This doesn’t exist in the official API; if it does, it’s not official.
• Price per message without a rate table, or "credit packages" without specifying how many messages each credit covers per category.
• Guaranteed green badge. Verification and display name depend on Meta, not the provider.
• Long contract as a condition to activate the number. High exit costs usually compensate for a service that can’t retain clients on its own.

Can you go directly to Meta without a provider?

Technically, yes. The Cloud API is open to developers: the company creates an app on Meta, registers the number, and integrates it with its own system. This makes sense in three situations:

• The company has a tech team to maintain the integration, handle errors, and keep up with Meta’s changes over the years.
• The use is purely technical, like sending verification codes or order notifications from an existing system, without human support.
• The volume is large enough that any margin per message outweighs the cost of developing and maintaining the system.

For those needing support with a team, funnel, and automation, going direct usually costs more: the company pays for development to rebuild what a tool already delivers and assumes maintenance alone. The useful question isn’t "provider or Meta," but "how much does it cost to build and maintain what I need to use every day."

How long does it take to get started

Another point worth asking is the path to the first message, because it varies little from provider to provider when the process is official:

• Connecting the number via embedded registration takes minutes, provided the person has admin access to the portfolio.
• Template approval usually happens quickly when the text follows category rules; poorly written templates are rejected and delay everything.
• Company verification depends on documentation and can take longer; it doesn’t prevent starting but affects limits and the displayed name.
• Sending limit starts low and increases as the number maintains good quality.

A provider promising "everything working tomorrow with unlimited sending" is promising something beyond their control.

API alone doesn’t serve anyone

A common mistake is contracting "the API" and later discovering it’s just plumbing: it receives and sends messages. For the operation to work, the company needs a tool on top of it, and that tool is what the team will use eight hours a day. Also evaluate:

• Shared inbox with ownership, status, and queue, so no conversation is ownerless.
• Template creation and tracking without needing to enter Meta’s manager for every adjustment.
• Automation that responds first and hands off to a person at the right time.
• Sales funnel and reports linked to conversations, not in a separate spreadsheet.
• Other channels in the same inbox, like Instagram Direct and Messenger, when customers write from multiple places.

Often the right decision is to choose the support tool first and only then check if it connects cleanly via the official API — not the other way around.

If you’re already with the wrong provider

Discovering the account is in the provider’s portfolio or that the margin per message is expensive is not the end. A number can be migrated between providers while keeping the same number, and the process is predictable when well planned:

• Confirm which portfolio the account is in and formally request transfer if it’s not yours.
• List approved templates and their text before leaving.
• Export the conversation history that the operation needs to keep.
• Choose a low-traffic time for the switch and notify the team.

The full step-by-step is at how to migrate a number between API providers.

Common mistakes in choosing

• Choosing based on monthly fee. The growing cost is per message, not the tool.
• Not reading the data contract. Customer conversations are personal data. The contract must specify where data is stored, who accesses it, how long it’s kept, and how it’s deleted when the company leaves — the company, not the provider, is responsible to the client under LGPD.
• Leaving only one person as portfolio administrator. If they leave the company, access goes with them.
• Signing without testing support. Send a technical question before closing and time the response.
• Separating support and sending across different providers without integration. The customer replies to the campaign and no one sees it.
• Confusing official with unofficial. Ask in writing if the connection is via Meta’s official API. Connection via common QR Code is not, and the risks are different — see Unofficial WhatsApp API: the real risks.

The API is the same for everyone. The provider is not. Choosing based on account ownership, total cost, and the tool the team will use avoids paying twice: once when contracting and again when leaving.