Key takeaway
Mass email marketing is one message sent to a large, defined audience as a single job. The size that makes it mass isn't yours to choose: Gmail treats more than 5,000 messages a day to Gmail accounts as bulk sending, and a Workspace mailbox caps out at 2,000 a day whatever you do. The message and the rate are easy to fix in an afternoon. Where the list came from is the one input I've never seen anybody fix afterwards.
What mass email marketing actually means.
The definition is narrow: one message, one large and defined audience, one job, fired on a schedule you set rather than in response to something an individual recipient did. That's the whole of it. It's also why people get the threshold wrong, because "mass" reads as a size you pick once the list starts to feel big. It isn't a size you pick. It's a line the receiving side draws, and once you're over it your mail has to carry things it didn't have to carry before. Google's sender guidelines treat anyone sending more than 5,000 messages a day to Gmail accounts as a bulk sender, and have done since 1 February 2024.
So the useful version of the definition has a number in it. Under 5,000 a day to Gmail addresses, you're sending email. Over it you're operating a sending program, and the difference isn't ambition or budget, it's the set of rules that now applies to everything you put out. I've watched teams cross that line in the middle of a launch week without anyone noticing, because nothing in the tool they were using told them they had.
What the definition leaves silent is the part that matters. It describes the send, the audience, the schedule, and the message, and it says nothing at all about who's allowed to receive it. That's the half that decides whether the send was worth making.
Time is the main reason businesses look for bulk sending tools. My previous e-commerce company hit this exact wall. I put it bluntly on In the Ring with SUMO Heavy: "our biggest constraint was actually time." You know the messages matter. You just run out of hours to build them manually.
The threshold that makes you a bulk sender.
Above 5,000 messages a day to Gmail accounts, four things stop being good practice and start being requirements. SPF and DKIM both configured. A DMARC record on the sending domain, where an enforcement policy of p=none is enough to satisfy it. The From: domain aligned with either the SPF domain or the DKIM domain. And, on marketing and subscribed mail, one-click unsubscribe alongside a clearly visible unsubscribe link in the body. Yahoo asks for the same authentication and adds a clock to it: unsubscribes honoured within two days, per its sender best practices.
The number that actually binds is the complaint rate. Both providers put it below 0.30%, and Yahoo measures it against mail delivered to the inbox rather than against everything you sent, which makes it stricter than it looks. Three people in a thousand reaching for the spam button is your ceiling. It's also the only campaign metric where crossing a line changes what happens to the next send instead of just reading badly in a report. One-click unsubscribe exists to keep people off that button, and mechanically it's RFC 8058: a List-Unsubscribe-Post header that lets the mailbox provider fire an HTTPS POST on the reader's behalf, with no confirmation step and a valid DKIM signature required. The reader taps once and they're gone.
Meeting every requirement here gets your mail accepted, not read. The threshold is a floor, and clearing it says nothing about whether anybody wanted the message. It's cumulative too, in a way the checklists miss. A domain crossing 5,000 a day for the first time isn't judged from that morning onwards, it's judged on everything it sent before it crossed, which is why the authentication work is worth doing on the day you send your first hundred.
Your mailbox isn't a mass email tool.
A mailbox is built to send one message to a few people. A sending platform is built to send one message to a lot of people. That sounds like a difference of degree and it's really a difference of design, and the first place it shows up is a published limit. There's a whole genre of how-to that teaches mass sending straight out of Gmail, and some of those guides open by suggesting you create a few more Gmail accounts to get around the ceiling. The providers publish that ceiling, and they're specific about it.
| System | Daily limit | Per-message limit | Notes |
|---|---|---|---|
| Gmail in Google Workspace | 2,000 messages a day, of which 1,500 can be mail merge | 2,000 recipients per message, no more than 500 of them external | |
| The Gmail API and SMTP relay | 500 recipients per message through the API, 100 per message through SMTP relay from a POP or IMAP client | Google's published limits | |
| Exchange Online | 10,000 recipients a day | Up to 1,000 recipients per message, 30 messages a minute | Microsoft says isn't suited to bulk mailing |
Microsoft goes further than the numbers and tells customers who need to send legitimate bulk commercial email to use a third-party provider that specializes in it. The limits are only the visible half of why. A mailbox has no suppression list applied at send time, no complaint feedback loop, no per-domain throttling, and no way to hold a send when the email bounce rate moves. Those absences cost more than the ceiling does, and you find out about them later than you'd like.
The list is what breaks at volume.
A mass send has three inputs: a message, a rate, and a list. Two of them are yours to fix in an afternoon. Where the addresses came from is the one you can't fix after the fact, and the industry position on the worst version of it isn't ambiguous. Appending email addresses to customer records, where the owner never gave you the address and never consented to it, is named a direct violation of core M3AAWG values, and a route to complaints, rejections, and a domain nobody trusts.
Volume doesn't create a list problem. It prices one. A 2% bad-address rate on a send to 500 people is ten bounces nobody notices, and the same 2% on a list of 200,000 is four thousand bounces, paid for out of the sending reputation the next campaign depends on. The complaint threshold behaves identically, because 0.30% is a rate and not a count: the proportion that was survivable at small scale is the proportion that gets you filtered at large scale. That's the case for confirmed opt-in, where the recipient takes an affirmative action before anything else is sent. It sits at the top of the consent ladder, and it's the only acquisition method whose cost falls as volume rises.
Hygiene has a ceiling, and it's acquisition. Suppressing hard bounces and sunsetting addresses nobody has opened in a year keeps a good list good, and it won't turn a bought list into one. Nothing downstream fixes that either: honouring an unsubscribe on a weekly job at volume is a tidy-looking way to collect complaints, because the person who clicked on Monday gets Tuesday's send regardless.
Sending it, in six steps.
Six steps, and the first one isn't "pick a platform". Most guides on this subject open with a shortlist of email blast platforms because that's where the vendor wants you to start, and no platform fixes a domain that was never authenticated or a list nobody consented to. The order matters more than any single item in it.
- Authenticate the sending domain first. SPF, DKIM, and DMARC on the domain the mail is from, before a single message leaves. Email authentication is a DNS change and a wait, not a project.
- Ramp a new domain or IP instead of opening at full volume. Email domain warmup is a schedule you keep rather than a setting you flip.
- Define the audience as a segment that recomputes. Email segmentation evaluated when the message is built is a different thing from a CSV somebody exported on Tuesday.
- Apply suppression at send time. Unsubscribes, hard bounces, and complaints, checked as the message is built rather than as the list was.
- Send in batches at a rate you chose, and seed a holdback slice first. Releasing to the whole list at once is a decision you only get to make one time.
- Read complaint rate and bounce rate before open rate. Hold the next send if either moved, and read the same numbers independently in Google Postmaster Tools.
Those six are the send. Doing it again next month is the same six run as a job. Step two has an entire schedule behind it, and sending email at scale has an architecture behind it, and neither one belongs in the middle of a campaign plan.
Running a mass send from one command.
Look at what's actually in those steps. The audience is a query. The message is a template. The rate is a number. The suppression list is a table. The result is a stream of events. None of that inherently needs a person clicking through a builder, and at 200,000 recipients the clicking is the part that scales worst, because it's the only part that doesn't get cheaper the second time you do it.
Run your whole email stack from one agent command. Nitrosend is MCP-first, so every capability is an API endpoint and an MCP tool before it's a screen, and a mass send can be segmented, drafted, tested, rate-limited, and released from the same place the rest of the work already happens. Two things about that bear directly on volume. Contacts are unlimited on every plan, including Free, which counts because list size is what most platforms meter and a mass sender's list is the number that grows. BYO sending keys are available on Pro and above across Amazon SES, Resend, Postmark, Mailgun, and SendGrid, which counts because at volume the sending route is the largest line on the bill.
I've been building email platforms for over a decade, and the pattern holds: a product designed dashboard-first and then fitted with an assistant hands you a chat window bolted onto a builder. The tools reach whatever the screens already exposed, and the operations that end a job are the ones that stayed on a screen.
The honest half is that an agent doesn't decide whether the message was worth sending, and it doesn't make an unwanted list wanted. It takes the clicking away, and the clicking was never the work.
Where mass sends go wrong.
Four failures account for most of the bad sends I've watched, and not one of them is a copywriting problem. Two are decisions somebody made deliberately, under deadline pressure. The other two are what happens when nobody makes a decision at all.
- Mailing the whole list because segmenting 80,000 people felt like more work than segmenting 800.
- Buying, renting, or appending addresses to reach a number the campaign plan had already promised somebody.
- Releasing the entire send in one go so the first sign of trouble arrives after the last message has landed.
- Reading a bounce spike as a subject-line problem and rewriting copy while email deliverability keeps falling.
The first two are decisions. The second two are defaults, and the defaults are the ones a platform should have taken off you. I'd add a fifth from experience: not watching the complaint rate until it's already past the ceiling. The automatic flag I ran at my last company sat at 0.5%, five times Google's ceiling, and it still wasn't catching problems early enough.
A mass send is a segment, a template, a rate, and a suppression list. Running one shouldn't mean an afternoon of clicking through a builder and hoping the complaint rate holds. Nitrosend specializes in email, and all four of those are an endpoint and an MCP tool before they're a screen. Start free with 8,000 emails to begin with, then 500 a month, unlimited contacts, and full MCP, API, and CLI access with no card.
Sources
- Google, Email sender guidelines: the 5,000-messages-a-day bulk sender threshold and the 1 February 2024 start date, the SPF, DKIM, and DMARC requirements with alignment, the 0.30% spam rate, and one-click unsubscribe.
- Yahoo, Sender best practices: the same authentication requirements, a spam rate below 0.3% measured on mail delivered to the inbox, and unsubscribes honoured within two days.
- Google Workspace, Gmail sending limits: 2,000 messages a day, 1,500 for mail merge, 2,000 recipients per message with 500 external, 500 per message via the Gmail API, and 100 via SMTP relay.
- Microsoft, Exchange Online limits: 10,000 recipients a day, up to 1,000 per message, 30 messages a minute, and the statement that Exchange Online isn't suited to bulk-mailing scenarios.
- M3AAWG, Sender Best Common Practices 3.0: email appending as a direct violation of core values, and confirmed opt-in as the highest consent standard.
- IETF, RFC 8058: the one-click unsubscribe standard, defining the List-Unsubscribe-Post header, the HTTPS POST with no confirmation step, and the DKIM signature requirement.
Common questions
There's a published number rather than an adjective. Gmail treats anyone sending more than 5,000 messages a day to Gmail accounts as a bulk sender, and above that line it requires SPF, DKIM, and DMARC on the sending domain, an aligned From: header, one-click unsubscribe on marketing mail, and a spam complaint rate below 0.30%. Yahoo asks for the same authentication without publishing a volume. Below that you're sending email; above it you're running a sending program that gets judged as one.
Only up to a point, and the providers publish where the point is. Gmail in Google Workspace allows 2,000 messages a day, 1,500 of them for mail merge, with 2,000 recipients per message and no more than 500 external. Exchange Online allows 10,000 recipients a day and 30 messages a minute, and Microsoft says plainly it isn't suited to bulk mailing and points customers sending legitimate bulk commercial email at a provider that specializes in it.
It can be, and the answer is about consent and honesty rather than about volume. Sending a lot of mail to people who asked for it is a programme. Sending the same mail to people who didn't is how a domain gets filtered, and no amount of hygiene downstream repairs it. Permitted is a floor, not a strategy: a send you're entitled to make can still be one that collects complaints.
Probably not. A dedicated IP below roughly 200,000 messages a month is a mistake in our view, because there isn't enough consistent volume flowing through it to build a reputation on, and a quiet IP looks worse to a receiver than a shared pool with steady traffic through it. Reputation is judged on the domain as well as the IP, and the domain is yours either way. Get authentication, list quality, and a consistent sending rhythm right first.
Usually one of three things, in this order. Authentication: SPF, DKIM, and DMARC missing or misaligned on the sending domain. Complaints: a spam rate creeping toward the 0.30% ceiling that Gmail and Yahoo both enforce. Or the list itself, where a chunk of the addresses never asked for the mail, so bounces and complaints arrive together. Subject lines and spam-word checkers are the last place to look and the first place most people start.
It depends on what your platform meters, and platforms meter one of two things: the size of your list, or the number of emails you send. If your list is much larger than your monthly volume, contact-based pricing is the expensive model for you, and if you send heavily to a small list it's the cheap one. At volume the sending route itself becomes the largest line on the bill. Nitrosend's free tier is 8,000 emails to start, then 500 a month, with unlimited contacts and full MCP, API, and CLI access.
Nothing mechanical. They're the same send under two names, and the difference is connotation: a blast implies one message going to everybody with no segmentation behind it, which is exactly the sending pattern that pushes complaint rates up. If you use the words interchangeably, nobody will misunderstand you. If you send the way the word blast implies, the receiving side will notice before your readers do.
All of them. Segmenting the audience, building the message, setting the rate, applying suppression, sending a test, and releasing the batches are each a call, so a script, an agent, or the command line can run the whole job. Nitrosend puts every capability behind an API endpoint and an MCP tool, and the screens sit on top of those same calls rather than in front of them. The results come back as data too, so you can read the complaint rate wherever you already work.