Booking confirmations needed to go out on WhatsApp. That was never in question. In India, an email confirmation is half a confirmation — people live in WhatsApp. A booking that doesn't ping their WhatsApp feels, to a lot of users, like it didn't really happen. So a confirmation message on WhatsApp wasn't a nice-to-have. It was table stakes.
The obvious choice was Twilio. It's the default. It's what everyone reaches for, it's what every tutorial uses, it has a clean SDK, great docs, and a reputation built over a decade. I'd used it before. So I did the sensible-looking thing and started wiring Twilio in as the channel for WhatsApp confirmations.
And then I ran into the wall that the tutorials don't put on the first page.
Sending a WhatsApp message through Twilio — a real, business-initiated message to a customer — is not "call an API and you're done." There's a whole compliance machine behind it. You need an approved WhatsApp Business sender. You need message templates pre-approved before you're allowed to send them. There's a verification process, a waiting game, an approval queue, and a pile of setup that exists for good reasons but stands directly between a solo builder and the simple act of sending "your booking is confirmed."
I sat with the timeline and did the math against my own promise. I'd told my friends, over ice cream, that I was launching this month. The Twilio WhatsApp path had an approval-shaped hole in the middle of it that I did not control and could not rush. I could do everything right and still be sitting on my hands waiting for a verification to clear while my self-imposed deadline slid past. The blocker wasn't my code. It was a queue I wasn't at the front of.
So I made the call that's hard for an engineer to make: I ripped out the Twilio integration I'd already started. Working-in-progress code, deleted on purpose. Again. (This is becoming a theme in this series — me lovingly building a thing and then tearing it out once reality shows up. Build it wrong first, apparently, applies to integrations too.)
I switched to AiSensy — a WhatsApp Business platform built specifically for the Indian market, designed to get businesses sending on WhatsApp faster, with the template and approval flow handled in a way that fit my timeline and my context. Right tool for where I'm building and who I'm building for, not the most famous tool in the room.
Here's the lesson I keep relearning, now stamped a little deeper: the best-known tool is not automatically the right tool. Twilio is excellent. Twilio is also built for a global, enterprise-shaped world, and I'm one person trying to send a booking confirmation to an Indian creator's client before the end of the month. "Most powerful" and "right for my exact situation" are different questions, and confusing them costs you days.
And there's a quieter discipline underneath this whole episode: protect the launch date. Every time something threatens the date, the question isn't "can I make this work eventually." It's "can I make this work by then." Twilio could work eventually. AiSensy could work by then. By-then wins when you've made a promise.
WhatsApp confirmations are flowing now. The date is still alive.