Best email marketing software 2026 — 7 tools compared

Choosing an email tool usually comes down to the same question: is a simple newsletter tool enough for me, or do I really need automation. The answer decides the price — but so does what the service bills by: list size or the number of emails sent.

The short answer

• Free plan with no sending limit

  Systeme.io

  2,000 contacts and unlimited sending without paying anything; measured by subscribers the most generous is EmailOctopus with its 2,500.

  Start free — 2,000 contacts (ad)

• Cheapest for a growing list

  EmailOctopus

  2,500 subscribers free, and after that €16/month where Brevo asks €21, MailerLite €29 and Mailchimp €39.11.

  Start free — 2,500 subscribers (ad)

• Cheapest for a large list

  Brevo

  A 10,000-contact list costs €28/month, while MailerLite asks €79 and ActiveCampaign €149.

  Start free — 300 emails/day

Comparison table

Prices are monthly prices excluding VAT at three list sizes, read from the services’ own price lists on 3 September 2026. For each we took the cheapest paid plan that supports that list size.

← scroll →

Prices exclude VAT, billed monthly unless stated otherwise.(ad) = ad link

• The price is the monthly price of the cheapest paid plan at that list size. The Free plan column shows when you do not have to pay.
• EmailOctopus and ActiveCampaign: the price list only shows the monthly price on annual billing. Monthly billing costs more.
• GetResponse and ActiveCampaign: the smallest list sold is 1,000 contacts, so the 500-contact column shows that price.

Two pricing models — choose the model first, then the tool

The biggest single price difference in the comparison does not come from features but from what the service uses as its meter.

• Contact-based. MailerLite, Mailchimp, ActiveCampaign, GetResponse and EmailOctopus bill by list size. Send once a month or ten times — the price is the same. As the list grows the price rises in steps, and at 10,000 names the steps are already big.
• Volume-based. Brevo bills for emails. A 10,000-name list fits two newsletters a month in the 20,000-email package, which costs €28 — the same list costs €149 at ActiveCampaign. If, on the other hand, you send to the same list weekly, the calculation turns the other way.

A practical rule of thumb: work out how many emails you send a month (subscribers × sends). If the figure is below about twice the list size, the volume-based model is cheaper.

How contacts are counted — the rule the price list does not tell you

The price list says what a thousand contacts cost. It does not say what counts as a contact — and that is the biggest hidden cost in the comparison. The rule is on each service’s own help pages, not on the price list, and it differs between services more than the price itself does.

The difference is money directly, because unsubscribes accumulate on every list. At Mailchimp an unsubscribed contact counts toward the monthly limit until it is archived separately; at MailerLite and EmailOctopus it does not. The same list can therefore be a different size on two services’ meters, even though the subscribers are the same.

← scroll →

The rules are read from the providers’ own help pages. “Not read” means the rule could not be opened in the original source — we do not guess it in either direction.

• Systeme.io. The plan limits are contact counts (free plan 2,000, Startup 5,000, Webinar 10,000). All nine contact-management articles in the help centre were read on 5 September 2026: none of them says which contact statuses count toward the plan limit. The guides advise cleaning the list, but the counting rule is not published. So we claim nothing either way until it can be read or has been asked from support.
• EmailOctopus. Unsubscribed contacts do not count toward the limit, nor do hard bounces and spam complaints, which are marked as unsubscribed automatically. Contacts awaiting double opt-in do not count either. One trap: the same address is counted separately for every list — an address on six lists is six contacts.
• Brevo. The price is set by monthly sending volume rather than contact count, so the question of counting unsubscribed contacts does not apply to Brevo the way it does to the others. The flip side is that the same list costs more the more often you write to it.
• MailerLite. The limit counts subscribers in Active status; unsubscribed and bounced do not count. One trap in the billing period: the period is counted cumulatively, so an address that was active at any point in the period counts toward the limit until the end of the period, even if you delete it immediately. Cleaning the list shows on the bill only in the next period.
• Mailchimp. An unsubscribed contact counts toward the monthly limit, as do non-subscribers who ended up in the account by other routes. Archived, cleaned and deleted contacts do not count, so lowering the bill takes a separate step: the contacts have to be archived, not just allowed to unsubscribe.
• GetResponse. Billing is set by two figures: the month’s highest number of active contacts and the total number of contacts added during the month, regardless of what later happens to them. The same address on several lists is counted several times. Exceeding the limit is not a one-off charge but recurs every billing period, on an annual contract too, and above 100,000 contacts the charge is $4 for every thousand started.
• ActiveCampaign. The rule depends on the account creation date. ActiveCampaign’s own help article gives two answers: on accounts created on or after 3 November 2025, all contacts count toward the limit regardless of status, i.e. unsubscribed, unconfirmed and bounced too; on accounts created on or before 2 November 2025, only active ones. A new account is therefore in the same group as Mailchimp in this comparison. The help page opens only in a browser (HTTP 403); the text was read through the help centre’s own API on 5 September 2026.

Two practical conclusions. When the same address is counted separately for every list, keep one list and separate readers with groups or tags. And because cleaning shows on the bill with a delay, clean the list before the billing date, not after the limit has already been exceeded.

What the research says about sending frequency

Estimating sending volume is worth doing before choosing a tool, because it decides which pricing model is cheaper. There is published research on it too, and it says something other than “send as often as you can”.

A study published in the Journal of Marketing Research (Zhang, Kumar and Cosguner 2017) modelled a US hardware retail chain’s email programme at the customer level. The effect of sending volume on returns turned out to be non-linear: both too sparse and too frequent sending reduced customer lifetime value. At that company the optimum varied between five and fourteen emails a month per customer and averaged seven. According to the American Marketing Association’s summary, dropping to four would have cost the company 32 per cent of lifetime value and raising to ten 16 per cent.

The researchers themselves say the figures are one company’s figures and should not be generalised. That is exactly why the shape, not the number, is what to take from this: the right sending frequency is a range, and money is lost on both sides of it. For choosing a tool this means two things. If your sending frequency lands in a range where emails pile up relative to list size, the contact-based model is cheaper. And if you plan to test frequency, choose a tool where changing frequency does not change the bill every month — otherwise the price of the test gets mixed up with its result.

The second measured point concerns what to demand from the tool. In a randomised field experiment across three companies published in Marketing Science (Sahni, Wheeler and Chintagunta 2018), merely adding the recipient’s name to the subject line raised the probability of opening by 20 per cent (from 9.05 per cent to 10.80), increased leads by 31 per cent and reduced unsubscribes by 17 per cent. The scale of the experiment is notable: the emails went to millions of recipients. For choosing a tool this is a cheap lever compared with switching services — the experiment needs one field in the subject line, not a pricier plan. Check your own tool’s help pages for what the field is called there.

The best choice by situation

• First newsletter, list under 2,000 → Systeme.io: the free plan goes a long way and sending is not capped. Try it free (ad)
• Growing list, small budget → EmailOctopus: free up to 2,500 subscribers and the cheapest paid plan after that. Try it free (ad)
• Large list, newsletter sent rarely → Brevo: the price follows sending volume, so a 10,000-name list costs a fraction of the others. Try it free
• Multi-step automation → ActiveCampaign: conditional content from the Plus plan, scoring from the Pro plan and a sales pipeline in the same view — priced accordingly. Try it free
• Courses or sales funnels → Systeme.io: everything in one price, no separate course platform. Try it free (ad)
• Webinars as part of marketing → GetResponse: the webinar tool is built in, no extra fee. Try it free

The tools one by one

Free plans shrank — what that means for a beginner

Only a few years ago a free plan meant a thousand or two thousand contacts at almost all of them. Now the range has torn open: Mailchimp’s free plan holds 250 contacts and 500 emails a month, MailerLite’s 250 contacts and 2,500 emails, while EmailOctopus gives 2,500 subscribers and Systeme.io 2,000. GetResponse and ActiveCampaign offer no free plan at all, only a 14-day trial.

For a beginner this means one thing: check the free plan limit before building the list, not after. Moving a list from one service to another at 300 names is an hour’s job; at 3,000 names it is a weekend.

Deliverability: what can be read from the help pages

We do not test it ourselves. Below are two things: Tooltester’s 2017–2024 series as history — it ended in January 2024, and the source no longer stands behind its method — and eight points from the services’ own help pages, which are the current basis for comparison.

← scroll →

Tooltester ended the series with the January 2024 round and explained why in August 2026: seed lists proved unreliable and providers started optimising for them. The figures are therefore history, not a measurement of current deliverability.

Source: Tooltester deliverability test, 12 rounds 8/2017–1/2024. About 80 seed addresses, 28 providers, shared IP addresses. Average across all platforms in the latest round 83.1%. Average and range calculated from the published rounds.

Deliverability features from the help docs

Eight points read from each provider’s own help pages on 13 September 2026. This replaces percentages as a buying basis: a feature can be checked and dated, a seed-list result no longer can. Every cell links to the article it was read from.

← scroll →

Yes = the help page describes the feature. No = the help page says it is absent or points to an external tool. — = not found in the help pages; that does not mean the feature is absent.

Read from the help pages, the basics are there at all seven: authentication records from the app, bounces and complaints removed automatically, purchased lists banned. The differences are in the tools. Brevo is the only one that shows a deliverability score in the app. List cleaning is in the app at four — MailerLite (“Clean up inactive”), Mailchimp (Omnivore checks imports), GetResponse (List hygiene) and ActiveCampaign (Engagement Management) — while Brevo, EmailOctopus and Systeme.io point you to manual work. A dedicated IP and a deliverability team are offered only to large senders or on the most expensive plan, and not at all at EmailOctopus and Systeme.io. As history: in Tooltester’s series Mailchimp was the steadiest of the five we compare and Brevo the most volatile, but the series has ended.

What Gmail and Yahoo require of senders

Part of deliverability is not a feature of the tool but the sender’s own responsibility, and that part changed in February 2024. Google and Yahoo published their requirements in two parts. The basic ones apply to every sender: SPF or DKIM, and a spam rate below 0.30%. Senders sending more than 5,000 emails a day must also have both SPF and DKIM, DMARC and one-click unsubscribe. The requirements can be read in both companies’ own guides, and they are unambiguous in a way that is rare in email.

• SPF, DKIM and DMARC under your own domain. Google’s guide additionally requires that the domain in the sender’s From header matches either the SPF or the DKIM domain, and that the DKIM key is at least 1,024 bits.
• One-click unsubscribe. Marketing emails must support it in accordance with RFC 8058: the recipient’s click is enough, and no further confirmation may be required. Yahoo’s guide also sets a deadline — the unsubscribe must be processed within two days.
• Spam complaint rate below 0.30 per cent. Google states the same figure twice: you must stay below it, and 0.30 per cent or more is a line you must not reach.
• Emails are sent over a TLS connection and formatted according to RFC 5322.

Sources: Google’s sender guidelines and Yahoo’s sender guidelines, read 5 September 2026.

In 2025 a third joined. Microsoft announced in its own blog post (2 April 2025, updated 29 April) the same requirements for Outlook.com’s consumer service, i.e. hotmail.com, live.com and outlook.com addresses, when more than 5,000 emails a day leave the sender’s domain: SPF and DKIM must pass, and a DMARC record must be in place at least at the p=none level, aligned with one of them. From 5 May 2025 an email that failed the requirements was routed to the spam folder, and Microsoft announced it would later reject such emails outright with error 550 5.7.515. A working unsubscribe link is a recommendation on Microsoft’s list, not a requirement — at Google and Yahoo it is a requirement.

For the reader this leads to a simple conclusion: the three largest mailbox providers now require practically the same thing, and the requirement falls on the sender, not the tool. The same authentication satisfies all three, so it is done once.

For a small sender one concrete thing follows: you need your own domain before the first newsletter, and you need access to its DNS records. A free email service address will not do as a sender, because its domain cannot be authenticated. MailerLite’s own guide says it plainly: “Emails sent from free domains can’t be authenticated, which can significantly affect your email deliverability and sender reputation.”

This is good news for a tool comparison. Authentication is done the same way at every service — the service gives you DNS records, you add them to your domain’s management — so it is not a reason to choose a service but a setup step that belongs to every one of them. It is also why, when comparing free plans, it is worth checking whether the service lets you use your own domain for free.

GDPR: what a European buyer should check

US comparison sites almost always skip this section, although for a buyer in the EU it is often decisive. Three things are worth going through before you move your list to a new tool:

• Data location. This is where the common assumption goes wrong. We went through every service’s own privacy policy on 3 September 2026, and five of seven store the list data in the EU: Brevo (France), GetResponse (Poland), Systeme.io (Irish ITACWT Limited, AWS Ireland), MailerLite (for EEA customers Irish MailerLite Limited, data centre in the EU) and EmailOctopus (British company, lists in AWS’s Irish data centre). The exceptions are Mailchimp and ActiveCampaign, whose transfers rest on the Data Privacy Framework and standard clauses — lawful, but worth knowing and recording in your own privacy documentation.
• Demonstrating consent. The tool has to store when and from which form the subscriber joined. We have not checked this at account level for every tool compared and therefore do not claim it for all of them — look it up in your own tool once, now.
• The right to erasure. When a subscriber asks for their data to be deleted, it has to succeed completely, not just as removal from the list.

A light practical rule of thumb: if you cannot answer these three questions about your tool, choose an EU provider. Consent, the duty to demonstrate it and data location tool by tool are covered separately: email marketing and GDPR.

How to land on the right one without a long comparison

If the list is under 2,000 names and you do not need automation, start on a free plan — Systeme.io or EmailOctopus. If the list is large but the newsletter goes out once or twice a month, Brevo is clearly the cheapest. If you build automations with conditions and branches, ActiveCampaign is worth its price; if you do not, it is the most expensive way in the comparison to send a newsletter.

Switching later is laborious but not impossible, so do not get stuck on the choice for months. The most common mistake is not the wrong tool but never sending the newsletter at all.

If you already know your use case, the choice has been made for you: email tool for an online shop and email tool for consultants and solo businesses.