How to Choose an Email Marketing Platform (What Actually Matters)

"Best email marketing platform" lists rank features you'll mostly never use, and skip the two things that actually decide whether email works for you: whether your emails reach the inbox, and what the platform will cost as your list grows. Nearly all of them charge by the number of subscribers, so the cheap plan you pick today can quietly become an expensive one at 20,000 contacts. This guide ignores the feature-count arms race and helps you pick by what matters — your use case, your projected cost, and deliverability — starting free.

Our full terms are on our disclaimer page.

Who this is for, and how we chose what to cover

This is for someone choosing or re-choosing an email platform for a business, newsletter or store in the US, with a list somewhere between zero and a few tens of thousands. Above that you have a deliverability team and different problems.

How we chose what to put in this guide, since that determines what you can trust it for:

• We cover the two criteria that are expensive to get wrong — per-subscriber cost at scale, and deliverability — because both are hard to reverse once you have automations, history and a sending reputation on a platform.
• We deliberately do not rank platforms or score their features. Feature parity in this category changes every quarter, and a ranking published today is misleading by spring.
• We name the platforms and link each one's own pricing page, because the alternative — telling you to "compare platforms" without saying which — is advice you cannot act on.
• We have not tested deliverability across platforms, and neither has anyone else in a way that is public, current and methodologically sound. Where we cannot know something, the guide says so rather than implying a winner.

Who this is not for: if you need only transactional email — receipts, password resets, no marketing — you may not need a marketing platform at all. See do you even need one.

What actually matters (more than the feature list)

Three things, in order:

1. Deliverability — the single most important job. A platform's real value is getting your email into the inbox, not the spam folder, and a meaningful share of entirely legitimate, permission-based marketing email doesn't make it — industry deliverability benchmarks have run in the region of 80–90% inbox placement, and they move year to year, so treat any specific figure you read as a snapshot rather than a constant. Litmus's guide to why deliverability matters covers the mechanics. What determines your side of it is sender reputation and email authentication (SPF, DKIM, DMARC) — set those up, and pick a platform with a good deliverability record, before you care about any fancy feature.
2. Owning your list. Your email list is an asset you control, unlike social-media followers you rent from an algorithm. That's the whole strategic point of email, and it's the same logic behind owning the platform you publish on rather than renting one.
3. Automation you'll actually use. You want to build a welcome series, a simple nurture sequence, and a re-engagement campaign without a developer — not 200 features you'll never touch.

The point: pick for deliverability, list ownership, and the two or three automations you'll really run — not for the longest feature list.

What you need before you sign up

Small list, but two items gate everything and people discover them halfway through setup:

If your domain sits with a host you chose on price alone, check now that it lets you add TXT and CNAME records — some budget panels make this harder than it should be.

The flagship: match the platform to your use case — and project the real cost

Step 1 — match to your use case. Platforms specialise, and the "best" platform for a newsletter is the wrong one for an online store:

Most offer free or low starter tiers, so you can begin at almost no cost and learn what you need.

The platforms themselves, named and unranked

Here are the names, grouped by the use case they are built around, alphabetically within each group, not ranked, not tested by us, and with no price asserted here. Each link goes to that platform's own pricing page, which is the only place the current per-subscriber number is reliable:

• Newsletters and publishing: beehiiv, Ghost, Kit (formerly ConvertKit), MailerLite, Substack
• General small-business marketing: Brevo, Mailchimp, MailerLite
• Ecommerce: Klaviyo, Mailchimp
• Sales process / CRM-led: Brevo, HubSpot

A platform appearing in two rows is doing two jobs adequately, not excelling at both. Note the structural differences that matter more than features: Substack and beehiiv are built for paid subscriptions and take a cut or a platform fee accordingly; Ghost is open-source and can be self-hosted, which changes the cost curve entirely; HubSpot's pricing is driven by marketing contacts and seats rather than subscribers alone, so it does not compare cleanly with the others on a per-subscriber basis.

Step 2 — project the cost at your future list size, not today's. This is the trap. Because platforms charge per subscriber, a plan that's free or $20/month at 1,000 contacts can be hundreds of dollars a month at 50,000 — and the gap between platforms widens as you grow (at a large list, one platform can cost several times another for the same thing). So before you commit, look up what your chosen platform costs at the list size you expect in a year or two, and compare a couple of options at that size — not at today's tiny list.

How to compare so the numbers mean something. Open two or three pricing pages and hold these constant: the same subscriber count, the same monthly send volume (several platforms cap sends as well as contacts), and the same features you will genuinely use — marketing automation, and landing pages only if you need them. Then add the two line items people forget: whether unsubscribed and inactive contacts still count toward your billed total (on several platforms they do, which is a quiet 10–20% surcharge on a poorly-maintained list), and whether transactional sending is billed separately.

Bottom line: start on a capable free tier that fits your use case, but choose with an eye on the per-subscriber price at scale — the cheapest platform now can be the most expensive later, and switching once you have automations and history is a hassle.

A worked example: the same list, two platforms, three years

Suppose you are starting a newsletter and you expect roughly 1,000 subscribers at the end of year one, 10,000 at the end of year two, and 50,000 at the end of year three. That is an ambitious but not fantastical trajectory for a niche publication that publishes consistently.

The platform prices below are hypothetical. They are not any platform's rates and are not intended to resemble any. They exist to show you the shape of the arithmetic, which is the transferable part — you will replace them with real figures from the pricing pages linked above.

The difference is $1,980 over three years, and $1,620 of it arrives in year three alone. That is the entire point: at 1,000 subscribers the two platforms look identical — both free — and the decision you make while they look identical is the one that costs you.

Now read it the other way, because this is where people over-correct. Platform B is cheaper. It is not therefore the answer. Ask what the $1,620 a year buys on Platform A: if it is an ecommerce integration that attributes revenue you would otherwise be guessing at, it may be trivially worth it. If it is a template gallery, it is not. The projection tells you the size of the question, not the answer to it — and its real value is that it makes you ask at year zero instead of year three.

What this example assumes, and what would change it. It assumes you actually hit those subscriber numbers — most lists do not, and a projection that never materialises makes you overpay early for a scale you never reach. It assumes flat pricing within a tier, when most platforms band pricing and you jump at thresholds. It ignores annual-billing discounts, which are common and material. And it ignores the largest variable of all: on several platforms, unengaged subscribers you never remove are billed the same as your best readers. A list hygiene habit is, quite literally, a pricing decision.

Bottom line: run this table with real numbers from two pricing pages before you sign up. It takes fifteen minutes and it is the highest-value fifteen minutes in this article.

Deliverability basics (the part that actually determines results)

An email nobody sees is worthless, so protect deliverability:

• Authenticate your domain (SPF, DKIM, DMARC) — most platforms walk you through it; do it. If your domain sits with a host you chose on price alone, check it supports the DNS records you'll need.
• Never buy an email list. Purchased lists tank your sender reputation, trigger spam complaints, and can put you on the wrong side of the law. US commercial email is governed by CAN-SPAM, and the FTC's compliance guide sets out the actual obligations — accurate headers, a working unsubscribe honoured promptly, a physical postal address, and no misleading subject lines. Penalties are per email, which is how a single bad send becomes expensive. Build your list from people who opted in.
• Keep the list clean. Remove inactive/bouncing addresses periodically; a list full of dead addresses and low engagement pushes all your mail toward spam.
• Watch engagement. Providers like Gmail judge you by opens and clicks; consistently ignored email gets filtered, so send wanted content to people who asked for it.

What to check: that your platform supports domain authentication and has a solid deliverability reputation — this matters more than any template gallery.

What Gmail and Yahoo now require of you

This is the part of the article that changed most since it was first written, and it changes the advice from "good practice" to "condition of sending".

Since February 2024, Google and Yahoo have enforced sender requirements for bulk senders. Both publish their own — Google's email sender guidelines and Yahoo's sender best practices — and both are the authority on their own thresholds, which they have changed before and may change again. Broadly, a bulk sender must:

• Authenticate with SPF, DKIM and DMARC. Not one of them; all three, with DMARC aligned to your sending domain.
• Offer one-click unsubscribe in the message headers, and honour it quickly. Note that this is a mailbox-provider requirement, not a CAN-SPAM one — CAN-SPAM requires a working unsubscribe mechanism, and Gmail and Yahoo additionally require this specific implementation.
• Keep spam complaints low — Google names a threshold below 0.3%, and advises staying well under it.
• Send from a domain with valid forward and reverse DNS, and use TLS.

Am I a "bulk sender"? The commonly cited threshold is around 5,000 messages a day to a given provider — but treat this as a bad planning assumption for two reasons. First, the count is per provider, and a single campaign to a modest list can cross it on a busy day. Second, and more importantly, the practices that are mandatory above the line are the ones that determine whether you reach the inbox below it. Nobody has ever regretted authenticating a domain early.

What this means practically: a reputable platform implements most of this for you and will not let you send until authentication is in place. What it cannot do for you is add the DNS records — that is yours, and it is why DNS access is on the prerequisites list above.

Setting up authentication, in order

Three records, and the order matters because DMARC depends on the first two working.

1. SPF — a DNS record listing who is allowed to send mail for your domain. Your platform gives you the value; you add it. One SPF record per domain, not several — multiple SPF records is the single most common misconfiguration and it fails silently.
2. DKIM — a cryptographic signature proving the message was not altered and genuinely came from your domain. Your platform generates the key and gives you a CNAME or TXT record to add.
3. DMARC — a policy telling mailbox providers what to do when SPF or DKIM fails, and, crucially, where to send reports. Start at p=none, which monitors without affecting delivery, read the reports for a few weeks, and only then tighten to quarantine or reject.

The exact record values are specific to your platform and domain, so they come from your platform's own setup flow — no article can give you a correct value. DMARC.org is the neutral reference for what the policy levels mean if you want to understand rather than copy.

Do not skip the p=none monitoring phase. Publishing a strict DMARC policy before you know every legitimate sender for your domain — your email platform, your CRM, your invoicing tool, your helpdesk — is how businesses silently stop delivering their own invoices.

How to know it's actually working

"Pick a platform with good deliverability" is not something a reader can check, and no independent current cross-platform dataset exists to check it against. What you can measure is your own inbox placement, which is the number that actually matters:

1. Read your DMARC reports. That is what the reporting address in your DMARC record is for. They tell you what is being sent as you and whether it authenticates. Free tools will parse the XML into something readable.
2. Use Google Postmaster Tools for the Gmail half of your list — it shows your domain reputation and spam-complaint rate against Google's own threshold. Gmail is usually the largest single slice of a consumer list, so this is the highest-value check available.
3. Send a seed test. Send a real campaign to your own addresses across Gmail, Yahoo, Outlook and an iCloud address, and look at where each one lands. Crude, free, and it catches a catastrophic problem immediately.
4. Watch the trend, not the number. A falling open rate across an entire list, with no change in content, is a deliverability signal before it is a content signal. Segment by mailbox provider if your platform allows it — a drop concentrated at one provider is diagnostic.

The test at 90 days: can you state your spam-complaint rate and name one mailbox provider where your placement is weakest? If not, you are sending blind, and the first sign of a problem will be a revenue number rather than a deliverability one.

The legal basics (CAN-SPAM)

U.S. commercial email is governed by the CAN-SPAM Act (FTC). The essentials: don't use deceptive subject lines or headers, identify the message as an ad where relevant, include a physical mailing address, and honor unsubscribe requests promptly (a working unsubscribe link is mandatory). It's not onerous — reputable platforms build most of it in — but the penalties for ignoring it are real, so don't.

What to check: your emails have a real unsubscribe link and your postal address, and you only email people who opted in.

One clarification worth having. CAN-SPAM does not require prior opt-in — it is an opt-out regime. Sending unsolicited commercial email to a purchased list can therefore be technically compliant and still be a serious mistake, because Gmail and Yahoo will judge you on complaint rates regardless of what the statute permits. The mailbox providers are the stricter regulator here, and they act faster. If you also have subscribers in the EU or UK, a genuine opt-in is a legal requirement there, not a preference.

The first 90 days

Sequence, not a schedule — how fast you move through it depends on how often you publish.

On cadence, since "be consistent" is meaningless on its own: pick a frequency you can hold for six months at your worst week, tell subscribers what it is in the welcome email, and hold it. The failure mode is not sending too often — it is sending irregularly, because a list that has not heard from you in four months treats your next email as unfamiliar, and so do the spam filters.

Do you even need a marketing platform?

Sometimes not, and the honest alternatives are worth knowing before you commit to a subscription:

When paid becomes worth it: the arithmetic is simple. If a paid tier costs $X a year and your list generates revenue, paid is justified when the additional revenue attributable to what the tier unlocks — automation, segmentation, an integration — exceeds $X. For most newsletters the honest answer in year one is not yet, and the free tier is not a compromise but the correct choice.

If you need to migrate later

You can move, and it is worth knowing exactly what it costs before you pick, because "you can always switch" is doing a lot of work in most platform advice.

What exports cleanly: your subscriber list, with sign-up dates and consent source if the old platform stored them. Take that CSV.

What does not: your automations rebuild from scratch; your templates rebuild; your historical performance data usually stays behind, so year-on-year comparisons break at the seam.

What you cannot export at all is the important one — your sending reputation. It is attached to the sending infrastructure, and on a new platform you start again. That is why a migration is normally done as a warm-up: send to your most engaged segment first, in increasing volume, over two to four weeks, before you send to the whole list. Send everything on day one and you will look exactly like a spammer who just bought a list.

Bottom line: switching is a fortnight of work and a temporary deliverability dip, not a catastrophe. It is also entirely avoidable by spending fifteen minutes on the cost projection now.

Common mistakes

• Choosing on feature count instead of deliverability and cost-at-scale.
• Ignoring the per-subscriber pricing and getting surprised by the bill as your list grows.
• Buying an email list. It damages deliverability and can break the law.
• Relying on social followers instead of building an owned list.
• Skipping domain authentication and wondering why open rates are low (your mail is in spam).
• Publishing more than one SPF record, which invalidates authentication silently.
• Jumping straight to a strict DMARC policy before monitoring, and blocking your own invoicing tool.
• Paying for subscribers who never open anything. On several platforms unengaged contacts are billed like everyone else — hygiene is a cost control, not housekeeping.
• Sending irregularly. A four-month gap damages both engagement and reputation more than a higher frequency would.
• Treating CAN-SPAM compliance as sufficient. It is the floor; the mailbox providers set the bar that actually governs whether you are delivered.

Putting it together

Email marketing lives or dies on deliverability and the list you own — not on which platform has the most features. So choose by use case (newsletter, ecommerce, or CRM-driven), start on a capable free tier, and — crucially — project the per-subscriber cost at the list size you're aiming for, because that's where platforms diverge and where you'll actually feel the price. Authenticate your domain, build your list honestly, follow CAN-SPAM, and keep the list clean. Do that and email becomes the highest-ROI channel you own; skip it and you'll pay for features that send your mail to spam.

Your next three moves, in order: (1) open two pricing pages from the table above and fill in the three-year projection with real numbers; (2) confirm you can edit your domain's DNS records before you sign up for anything; (3) authenticate all three records and set DMARC to p=none before your first send, not after.

Where to go from here

• If you are building the audience the list comes from, starting a publication that people actually subscribe to is the prior problem, and email is its most valuable output.
• If DNS access turns out to be the blocker, choosing a host that does not fight you covers what to look for.
• For the requirements that gate you rather than merely advise you, Google and Yahoo publish their own, and they are the only authoritative version.

Our full terms are on our disclaimer page.

FAQ

(Only questions the body doesn't fully answer.)

• Can I switch platforms later? Yes — you can export your list and move — but you may lose automation history and templates, and re-warming deliverability on a new sender takes care. It's doable but not free of friction, which is why projecting cost-at-scale up front matters.
• Is a free plan good enough to start? For most beginners, yes — the capable free tiers cover a small list, basic automation, and sending. Upgrade when your list size or automation needs genuinely outgrow it, not before.
• How do I estimate my real cost? Take the list size you realistically expect in 12–24 months and look up each platform's price at that tier — that number, not the free/intro price, is your real cost of ownership.
• Do open rates still matter? As a deliverability signal, yes — providers watch engagement — even though open tracking itself has become less precise. Focus on sending relevant email people actually open, which protects your inbox placement.
• What's a good open rate? Less useful a question than it was: privacy features that pre-load images inflate opens unevenly across providers, so cross-industry benchmarks are not comparable to your list. Track your own trend, and treat clicks and replies as the more honest engagement signals.
• Should I use a free-mailbox address as my sender? No. Sending bulk mail from a @gmail.com or @yahoo.com address fails authentication alignment and is explicitly discouraged by the providers. Use your own domain.
• Do I need a separate subdomain for marketing email? Useful once you send at volume — it isolates marketing reputation from the transactional mail your business depends on, so a bad campaign cannot stop your receipts arriving. Not necessary on a small list, and it requires its own warm-up.
• How long does authentication take to work? The records themselves take minutes to add and typically propagate within hours, occasionally up to a day. Reputation is different: it builds over weeks of consistent sending, which is why the warm-up exists.
• What if my list is old and I've not emailed it in a year? Do not send to all of it. Send to the most recently engaged slice first, in small volumes, and expect a high unsubscribe rate — that is the correct outcome, not a failure. Mailing a dormant list at full volume is one of the fastest ways to damage a new domain's reputation.