Key takeaway
Email marketing cost is three separate charges added together: the list you hold, the messages you send, and the hours somebody spends operating it. Published ranges bundle all three, which is why they run from nothing to thousands. Price them apart. Mailchimp Essentials is $13 a month at 500 contacts and $385 at 50,000, and that ladder is still the smallest of the three for most senders. The hours are the largest, and I've never seen a comparison spreadsheet count them.
What email marketing actually costs.
Email marketing cost is the monthly total of three things: the software that holds your list, the delivery of the messages themselves, and the labor of whoever runs the program. Published answers put that total anywhere between nothing and several thousand dollars a month, which is true and useless in the same breath.
The range is that wide because it's three unrelated bills added together, and each one scales on an input that has nothing to do with the other two. A $13 plan and a four-figure agency retainer are both honest answers to "what does email marketing cost". They're just not answers to the same question.
A range stops helping the moment you're holding a quote and want to know whether it's fair. The only thing that can tell you is knowing what each of the three charges should cost on its own, so that when one of them is out of line you can see which one it is.
For most senders under 50,000 contacts the platform subscription is the smallest of the three numbers, and it's the one everybody shops on. I've watched teams spend two weeks comparing $110 against $135 while paying somebody four figures a month to operate whichever one they picked.
The three charges hiding in one invoice.
Split any email bill and you get three charges. They scale on different inputs, they respond to different fixes, and only two of them ever appear on a statement.
- Holding a list: what you pay each month for contacts to sit in the database, whether or not you email them.
- Sending: what it costs to get a message out of the door and into a mailbox, priced per thousand at the infrastructure layer.
- Operating it: the hours somebody spends building, segmenting, cleaning, and monitoring, which is usually the largest of the three and the only one that never shows up on an invoice.
Almost every marketing platform prices the first two together and calls the result a subscription. That bundle is the thing worth noticing, because its two inputs move independently of each other. Your list grows every week whether you mail it or not. Your send volume is a decision somebody makes each month, and when the two are billed as one number the bill tracks the input you don't control.
It's also why the honest answer to "why did my bill go up" is so often "it didn't, your list did".
Advanced features justify a higher price tier when they directly increase sales. As of September 2020 we had "billions of emails" running through our infrastructure. I broke down the numbers on SaaS District to prove the value of personalization. Messages with data-driven product recommendations "make more than twice as much revenue" as standard newsletters. You pay for the intelligence layer because it generates a measurable return.
What the platform ladder charges for.
A platform ladder is a price per bracket of contacts, with a monthly send allowance attached as a multiple of that bracket. Every contact-priced platform works this way, so the shape is the same everywhere and only the steepness differs. Four rungs, lined up:
| Plan | Entry-tier price | Mid-tier price | Top-tier price | Send multiple |
|---|---|---|---|---|
| Mailchimp Essentials | $13/mo at 0 to 500 contacts | $110/mo at 5,001 to 10,000 | $385/mo at 40,001 to 50,000 | 10x the contact ceiling |
| Mailchimp Standard | $20/mo at 0 to 500 contacts | Not published in this comparison | $800/mo at 75,001 to 100,000 | 12x |
| Klaviyo Marketing module | $20/mo at 500 active profiles | $150/mo at 10,000 | $1,955/mo at 150,000 profiles | 10x the profile count |
| MailerLite Comfort | $12/mo at 500 subscribers | $49/mo at 5,000 | $179/mo at 25,000 | Not stated; priced on active subscribers only |
Read those left to right and the argument makes itself. Not one of those numbers is set by how much mail you send. They're set by how many people you keep, and the send allowance is derived from that afterwards as a multiple, which is why a sender whose volume never changes still watches the bill climb year on year.
The interesting figure was never the entry price either. It's what happens on the steep stretch. Klaviyo's slider steps in 500-profile increments between 10,000 and 13,500, and the price climbs from $150 to $325 across it. Mailchimp Essentials goes from $110 to $180 over the 10,000 to 15,000 step. So your bill is decided by which side of a bracket boundary your list happens to sit on, and list growth you never email crosses that boundary for you.
Software vendors pass their customer acquisition costs directly to you. Bidding on broad search terms is an expensive game for email platforms. I drew a hard line on SaaS District regarding our own ad spend. We are "not going to pay $19 for a click" to compete on generic keywords. We buy ads for misspellings and long tail phrases instead. Keeping our marketing budget low keeps your monthly invoice low.
What sending actually costs.
At the infrastructure layer a delivered message has a published unit price, and it's very small. Amazon SES charges $0.10 per 1,000 emails a la carte and $0.16 per 1,000 on its Essentials plan, so a million messages is $100 to $160 of delivery, and 10,000 messages is a dollar.
Set that against the ladder above and the postage turns out to be a rounding error. A sender at 150,000 profiles paying $1,955 a month has 1.5 million sends included, and 1.5 million sends is roughly $240 of delivery at Essentials rates. Everything else on that invoice is software, storage, support, and margin. That isn't an accusation. It's what a platform bill is for, and it's the number that tells you whether the next rung up is worth buying.
That unit price is a floor rather than a bill. Dedicated IPs, deliverability add-ons, and attachment data all sit on top of it, each on its own line: a standard dedicated IP is $24.95 per IP per month, and attachment data is $0.12 per GB sent. That IP line is the one worth questioning before you buy it: for Gmail specifically, domain reputation matters more than IP reputation, and a dedicated IP earns its keep only once you're sending past a million emails a month. Bare infrastructure also hands you no campaign builder, no automation, and no suppression model, which is the work the next section charges you for in hours instead.
The line nobody quotes: the hours.
The recurring operating work on an email program is list hygiene, authentication, suppression handling, and complaint monitoring. It isn't optional and it isn't a setting you switch on once, because the mailbox providers publish thresholds and then measure you against them continuously.
Google treats anyone sending 5,000 messages a day to Gmail addresses as a bulk sender and requires SPF, DKIM, and DMARC on the sending domain, a spam rate held below 0.30% in Postmaster Tools, and one-click unsubscribe on marketing mail. Yahoo publishes the same 0.3% ceiling and asks for unsubscribes to be honored within two days. None of that is a line item, and all of it is a reason a cheap plan gets expensive: a platform that can't hold suppression across two lists, or can't put a one-click header on a template, costs you the domain rather than the subscription.
A 0.3% ceiling is roughly three complaints in a thousand. In a decade of building email platforms I've never seen a team hold that line without one person owning it every week, and that person's hour is the most expensive input in the whole calculation. Price it honestly, at whatever an hour of their time really costs, and it dwarfs a $110 subscription before anybody has built a single campaign.
Those hours are almost entirely clicks: pull a segment, export a suppression list, check a spam rate, re-run a flow, paste the result somewhere a colleague can see it. Every existing platform was designed for a human doing exactly that, and an AI assistant bolted onto one still drives the same screens, so the clicks change hands rather than leave the bill. The labor line is the one charge of the three that responds to how a product is built rather than to how big a list has grown.
The true cost of any software includes the labor required to run it. I felt this acutely when running an e-commerce company before starting Nitrosend. Our biggest constraint was time. The reality I described on In the Ring with SUMO Heavy is that "there wasn't enough time for us to send all these emails" manually. You pay for a platform to buy back those lost hours.
How to price your own stack in four steps.
You need last month's invoice, your send volume, and a rough hourly rate for whoever runs the program.
- Count the contacts you actually emailed in the last 90 days, not the ones sitting in the database.
- Multiply your monthly send volume by $0.16 per 1,000 to get the delivery floor.
- Add the hours: campaigns built, lists cleaned, deliverability checked, times what an hour of that time costs.
- Subtract steps two and three from your current bill. Whatever is left is what you pay for software.
Interpreting the four numbers is where the value sits. If step one comes out much smaller than the list you're billed for, you're paying to store people rather than to reach them, and on some platforms that's fixable in an afternoon. Mailchimp counts subscribed, non-subscribed, and unsubscribed contacts toward the limit while cleaned and archived ones drop out, so archiving a dead segment can move you down a bracket without costing you a reader you were ever going to reach.
Step four is the software line, and it's the one everybody assumed they were shopping for. When it's large and step three is larger still, the platform isn't your problem and changing vendors won't fix it. Once step two starts to approach step four, you've reached the volume where buying delivery at cost and paying for the software separately becomes the cheaper shape.
A good platform makes it easy to track your return on investment. The first thing you see in our dashboard is the total revenue generated by your campaigns. I used a baseline on SaaS District of a store "getting 8% of our sales through email" to show how optimization works. You can only evaluate your software costs when you know exactly what the tool brings in.
Where the bill goes wrong.
Four failure modes account for most of the overpaying I see, and none of them is exotic.
- Paying to store contacts: the billed count includes people who unsubscribed on some plans, so a list nobody emails still sets the tier you're on.
- Buying a tier for one spike: a single seasonal campaign moves you up a bracket that then bills every month until somebody notices and moves it back down.
- Taking a dedicated IP too early: below roughly 200,000 messages a month it's a mistake in our view, and it's $24.95 an IP per month on SES before anyone spends weeks warming it.
- Stacking add-ons: deliverability insights, helpdesk seats, and hub modules are each priced on their own line, and none of them shows up in the headline plan price you compared.
Every one of those is a list-side or a tier-side charge. Not one of them is sending. That's the short answer to what email marketing costs: the postage is pennies, the software is a bracket you fell into, and the hours are the bill. Put a number against each of the three for your own account and you can finally say whether the quote in front of you is fair, which is more than any monthly range was ever going to do for you.
Price the three charges, then pick a platform.
Nitrosend is an AI-native email platform, built against that split. Every capability is an API endpoint and an MCP tool before it's a screen, so the operating work, the largest of the three charges, runs from one agent command instead of an afternoon of clicks.
The other two follow from it: contacts are unlimited on every plan, including Free, so the list you hold never sets the bill, and BYO sending keys on Pro and above put delivery on Amazon SES, Resend, Postmark, Mailgun, or SendGrid at cost. Nitrosend specializes in email, so if what you're buying on that ladder is product-feed sync with an ecommerce storefront, that isn't us, and paying the bracket price for it is the right call. Start free with 8,000 emails, then 500 a month.
Sources
- Google, Email sender guidelines: the 5,000-a-day bulk-sender threshold, the SPF, DKIM, and DMARC requirement, the 0.30% spam-rate ceiling, and one-click unsubscribe.
- Yahoo, Sender best practices: the 0.3% spam-rate ceiling and the requirement to honor unsubscribes within two days.
Common questions
Under 5,000 contacts, the software is usually the cheapest part: entry plans on the contact-priced platforms open around $12 to $20 a month, and delivery for a small program is under a dollar per 10,000 messages at infrastructure rates. What decides the real monthly figure is the third charge, which is the hours somebody spends building campaigns, cleaning the list, and watching the complaint rate. Price that at whatever an hour of their time costs and it's almost always larger than the subscription.
At the infrastructure layer, about a dollar. Amazon SES charges $0.10 per 1,000 emails a la carte and $0.16 per 1,000 on its Essentials plan, so 10,000 messages is $1.00 to $1.60 of delivery. On a marketing platform you don't buy those sends separately: the plan includes an allowance set as a multiple of your contact count, typically 10 to 12 times it, and the price you pay is for the contacts rather than for the mail.
Because contact-priced platforms bill on the size of the list, and the send allowance is derived from it afterwards as a multiple. Mailchimp Essentials is $110 a month at 5,001 to 10,000 contacts and $180 at 10,001 to 15,000; Klaviyo's ladder climbs from $150 at 10,000 active profiles to $325 at 13,500. Crossing a bracket boundary is what raises the bill, and signups you never email will cross it for you.
It's enough to start and rarely enough to run a program. Free tiers cluster around 250 contacts and a few hundred to a few thousand sends a month, which covers a first newsletter and stops the moment a signup form starts working. The useful test isn't the send cap, it's whether the free plan gives you the API and the automation you'll build on, because a free plan you have to leave in month two costs you the migration as well.
Agencies price four ways: an hourly rate, a fixed price per project, a monthly retainer, or a share of attributed revenue. No single figure is worth quoting, because the number turns entirely on which of those four you're buying and how many campaigns a month it covers. The useful move is to convert whatever you're quoted into hours, then compare it against what the same work costs in-house.
A share of budget is the wrong control, because it tells you nothing about whether the individual charges are fair. Price the three instead: what you pay to hold the list, what the delivery costs at infrastructure rates, and what the operating hours cost. A program where the software line is small and the labor line is enormous has a different problem to one where the two are reversed, and no percentage of budget distinguishes them.
Below roughly 200,000 messages a month, a dedicated IP is a mistake in our view: a low-volume IP sends too little traffic to build a reputation, and a shared pool with good neighbors will outperform it. A standard dedicated IP on Amazon SES is $24.95 per IP per month, and some platforms include one on a higher tier. The cost that matters isn't the line item, it's the warm-up: several weeks of ramped sending before the IP is worth anything.
Not usually as a line item, and almost always as hours. Authentication, one-click unsubscribe and suppression handling are built into every serious platform at no extra charge. What costs money is keeping inside the thresholds afterwards: a spam rate below 0.3% at Google and Yahoo, and unsubscribes actioned within Yahoo's two days. That's a weekly monitoring job somebody owns, not a setting.