Sendersy
Все статьи
SMTP и инфраструктура

Choosing an SMTP Relay Service: What Actually Matters

When to use an SMTP relay vs an API, what to look for, and how to migrate without breaking your existing app.

Igor Petrov
Инфраструктура, SMTP
8 июня 2026 г.6 мин чтения

An SMTP relay serviceaccepts mail from your application over the standard SMTP protocol and handles the genuinely hard parts — authentication, queuing, throttling, retries and sender reputation — before handing your message to the recipient's inbox provider. It is the lowest-friction way to make an existing application send email reliably, because almost every framework, language and content-management system already knows how to talk SMTP. You give it a host, a port and credentials, and mail starts flowing through a professionally managed pipe instead of a server you have to babysit yourself.

The alternative — running your own mail server, or worse, sending directly from your app server with a localsendmail — looks free until you count the cost. You inherit IP reputation from scratch, you fight blacklists, you build your own retry queue, you configure SPF, DKIM and DMARC by hand, and you become the person who gets paged at 3 a.m. when Gmail starts deferring your mail. A relay service exists so you never have to learn those lessons the expensive way.

What an SMTP relay actually does

When your app hands a message to the relay, a lot happens that you never see. The relay authenticates you, signs the message with DKIM under your domain, places it in a queue, and then delivers it to each recipient's mail server while respecting that provider's rate limits. If a server is temporarily busy and returns a soft failure, the relay backs off and retries on a schedule rather than giving up or hammering the server. If an address is dead, it records a hard bounce and suppresses it so you do not keep mailing a black hole. All of this protects the one asset that determines whether your mail reaches the inbox: your sender reputation.

SMTP relay vs HTTP API

Both an SMTP relay and an HTTP API can send the exact same email; they are two doors into the same building. The right choice depends on what you are connecting. Choose SMTP when you have legacy software, a CMS like WordPress, or an off-the-shelf tool that only accepts SMTP host, port, username and password. It is universal and requires no code. Choose an HTTP API when you are writing new code and want richer capabilities — structured JSON errors instead of cryptic numeric codes, templates, batch sends, scheduling, tagging and tighter integration with webhooks. The best part is that a quality provider gives you both against the same account, the same reputation and the same analytics, so you can use SMTP for the WordPress plugin and the API for your main application without splitting your sending identity.

Ports and encryption, briefly

SMTP submission happens on one of two ports. Port 587 with STARTTLS is the modern default: the connection starts in plain text and immediately upgrades to an encrypted TLS channel. Port 465 with implicit TLS encrypts from the very first byte and is equally valid. Avoid port 25 for submission — it is meant for server-to-server transfer and is widely blocked by hosting providers and residential ISPs precisely because spammers abused it. Whichever port you choose, the rule is absolute: always use TLS, so your username and password are never transmitted in the clear where they could be intercepted.

Sendersy

Send email that actually lands in the inbox

API and visual editor, SPF/DKIM/DMARC out of the box, analytics and warm IPs. Free tier — 200 emails/month, no card required.

What to look for in a relay

Beyond raw sending, the features that separate a serious relay from a hobby project are about safety and visibility. Look for TLS on the submission ports as a non-negotiable baseline. Insist on per-application credentials that you can create, scope and revoke independently, so a leaked key from one service does not compromise your entire sending and so you can rotate without downtime. Demand clear bounce and block feedback — a relay that tells you exactly why a message failed (mailbox full, domain not found, blocked for reputation) is worth far more than one that returns a generic error. And make sure SPF, DKIM and DMARCare handled for you, because authentication is the foundation of deliverability and you do not want it to be a manual project. Finally, check for a searchable delivery logand webhooks, because the day a customer says "I never got the email," you need to be able to answer with certainty.

Common SMTP errors and what they mean

SMTP speaks in numeric reply codes, and learning a few of them saves hours of confusion. A 250 means the message was accepted — success. A 421 or 450 is a temporary failure: the server is busy or rate-limiting you, and the relay will retry automatically, so it is usually nothing to worry about. A 550is a permanent rejection — the mailbox does not exist, or you have been blocked — and it should be treated as a hard bounce and suppressed. A 535 means authentication failed, which almost always points to a wrong username, password or a credential that was revoked. Seeing a flood of 421 deferrals usually means you are sending too fast for a particular provider; a good relay throttles to fix this for you, but it is a sign to check your sending pattern.

Migrating to a relay without downtime

Switching sending infrastructure feels risky, but a staged migration removes almost all of it. Point a single, low-volume, non-critical service at the new relay first — your password-reset flow is a good candidate because it is important enough to watch closely but low enough in volume to contain any surprise. Send through it for a day, watch the delivery log, and confirm that SPF, DKIM and DMARC all pass on real messages to Gmail and Outlook. Once you trust the new stream, move the rest of your services over one at a time, highest-volume last. Keep your old credentials and configuration in place until every stream is verified end to end, so you always have a rollback path. The whole migration can happen with zero customer-visible downtime if you take it one service at a time.

A simple, working configuration

host: smtp.sendersy.com
port: 587
secure: STARTTLS
user: your-smtp-username
pass: your-smtp-password

That is the entire configuration for most frameworks — drop those five values into your mailer settings and you are sending. Keep the password in an environment variable rather than in your code or a committed config file, and create a separate credential for each service so you can revoke one without disturbing the others.

Security and key hygiene

SMTP credentials are as powerful as an API key: anyone who has them can send mail as your brand. Treat them accordingly. Store them in a secrets manager or environment variables, never in source control. Use a distinct credential per application and per environment so you can trace and revoke precisely. Rotate them on a schedule and immediately if you ever suspect exposure. If a credential does leak, revoke it at once, issue a new one, and review your delivery log for any messages you did not send — fast revocation is the difference between a non-event and a reputation crisis.

Frequently asked questions

Can I use SMTP and the API at the same time? Yes, and you often should — SMTP for tools that only speak SMTP, the API for your own code. Both share one account and reputation.

Why are my messages being deferred? Deferrals (421/450) are usually temporary rate-limiting by the receiving provider. A relay retries automatically; persistent deferrals suggest you are sending faster than a provider allows or your reputation needs attention.

Do I still need SPF and DKIM if the relay handles sending? Yes — you publish the records the relay gives you in your own DNS so that mail sent through it aligns with your domain. The relay signs; you authorize.

Throttling: why slower is sometimes faster

One counterintuitive job of a relay is to deliberately slow you down. If you hand it a hundred thousand messages and it tries to deliver them all to Gmail in the same minute, Gmail interprets that burst as an attack and starts deferring or blocking — and now your whole send is stuck. A good relay paces delivery to each receiving provider according to that provider's tolerance, spreading a large send over time and backing off the instant it sees deferrals. The result is that a campaign which feels "slower" because it is throttled actually arrives more completely and more reliably than an unthrottled blast that trips every rate limit on the way out. Pacing is not a limitation; it is reputation protection, and it is one of the main reasons to use a managed relay instead of your own server.

Shared versus dedicated sending IPs

A relay sends from an IP address, and that IP carries reputation. Most senders should use a shared IP pool: their mail goes out alongside other reputable senders on IPs the provider actively manages and keeps warm. This is ideal for low-to-medium or bursty volume, because the pool already has an established reputation that no single sender could build alone. A dedicated IP only makes sense once you send a large, consistent volume — tens of thousands of messages a day — because a dedicated IP that sits mostly idle goes cold and its reputation suffers. The rule of thumb is simple: if you cannot keep a dedicated IP consistently busy, a well-managed shared pool will deliver better. Start shared and graduate to dedicated only when your volume genuinely justifies it.

Monitoring and the delivery log

The difference between a relay you trust and one you merely hope works is visibility. A proper relay gives you a searchable delivery log showing the status of every message — accepted, delivered, deferred, bounced or complained — along with the provider's response. When a customer insists they never received an email, you open the log, search their address, and see exactly what happened: delivered at 14:02, or bounced because the mailbox is full, or deferred and still retrying. Pair the log with webhooks so your application learns outcomes in real time, and connect Google Postmaster Tools so you can watch your reputation and spam rate from Gmail's side. Monitoring turns deliverability from a guessing game into something you can actually manage.

Get sending

An SMTP relay turns email from a fragile, self-managed liability into a reliable utility you barely think about. Sendersy gives you SMTP and an HTTP API side by side, with the same analytics, warm IP pools and automatic authentication behind both. Grab your SMTP credentials free and relay your first message today.

Поделиться:
Автор
Igor Petrov
Инфраструктура, SMTP

Отвечает за SMTP-релеи, очереди и мониторинг. Знает, почему письмо «зависло», ещё до того, как вы спросите.