When plain SMTP is the right call
Three situations make it the correct choice rather than a fallback.
Your host has no dedicated integration. Regional providers serve many markets better than the global names do, and most of them speak perfectly good SMTP.
Compliance or procurement dictates the path. Some organisations require mail to leave through specific infrastructure. SMTP is how you satisfy that without arguing about it.
You self-host deliberately. Postal, Haraka and similar stacks are real options for teams who want full control and have the expertise to run them.
The trade-off, stated plainly
The thing you give up is feedback quality.
An API-based provider hands you structured events — delivered, opened, clicked, bounced, complained — with identifiers you can trace back to a specific message. Plain SMTP gives you the response from the SMTP session, and whatever bounce mail arrives at your return-path afterwards.
SendNectar works fine either way. But suppression enforcement, reputation scoring, and per-recipient status are all downstream of event data, so a host with thin feedback gives you a thinner picture. Ask about this before you commit, not after.
Self-hosting: the honest version
Running your own MTA is legitimate and it is a real job. You take on IP warmup, blocklist monitoring, reverse DNS, TLS certificate management, queue tuning, and the entire relationship with mailbox providers when something goes wrong.
For most operators most of the time, a managed provider is the better trade — you are buying someone else’s deliverability team. Self-host when you have a specific reason and the capacity to support it, not to save money on a bill that is usually not the expensive part.
Notes for multi-brand senders
Rate limits are your safety net. Without a managed provider’s clear signals, per-connection hourly, daily and monthly caps plus max concurrency are what stop one brand’s send from overwhelming a shared relay and affecting the others.
Authenticate each brand separately — SPF, DKIM and DMARC per sending domain. There is no wizard here, so it is the step most likely to be done once for the first brand and quietly skipped for the rest.
Verify bounce handling per brand. If bounces from one brand’s return-path are not reaching SendNectar, its suppression enforcement is running blind for that brand while working correctly for the others — and that asymmetry is genuinely hard to notice.