GSendr didn’t start on Official WhatsApp Cloud API.
To ship WhatsApp campaigns quickly, we wired Evolution API — an unofficial stack. Connect a number, send, iterate. On a demo it looks fine. In production with real orgs, it didn’t hold.
What broke
Numbers got banned fast.
We weren’t blasting blindly. We put the usual safeguards in place:
- Working-hour send windows
- Daily send caps
- Per-message delays / jitter
- Throttling and careful pacing
It slowed the bleeding. It didn’t stop it.
Unofficial routes sit outside Meta’s business rules. You’re always one spam or quality signal away from a dead number. For a multi-tenant product, that isn’t a “ops will rotate SIMs” problem — it’s a customer-trust failure. An org that wakes up to a banned sender loses campaigns, reply context, and confidence in the platform.
The real cost
Every ban burned:
- Time reconnecting / replacing numbers
- Campaign history and operational continuity
- Trust with orgs that expected WhatsApp to just work
Limits and delays are table stakes. They don’t make an unofficial channel durable.
Speed of integration isn’t worth fragile deliverability.
What we moved to
We migrated GSendr to Meta’s Official WhatsApp Cloud API:
- Meta-approved templates and per-org template governance
- Bulk pipelines with retries, failure states, and delivery / read / failed reporting
- A real 24-hour customer-service window for free-form replies after the user responds
Boring on purpose: attributable outbound you can run across organizations without playing ban roulette.
Where that left us
Today GSendr runs Official Cloud API + email for roughly 10 organizations, with daily volume on the order of 10K+ WhatsApp and 10K+ emails. That volume only makes sense on a path Meta won’t randomly kill under you.
Takeaway
If you’re building WhatsApp into a product — especially multi-org SaaS — treat unofficial APIs as a prototype, not a foundation. We learned that the expensive way on Evolution API.
How the full platform is shaped (audiences, PDFs, reply desk, tenancy): How we built GSendr · Case study.