Sending a newsletter from WordPress sounds like it should be a solved problem. It is not — at least not in the sense that there is one obvious answer. There are three fundamentally different architectures, and the one that is right for a 400-subscriber craft blog is actively wrong for a 60,000-subscriber media property. The mistake most people make is choosing based on a comparison-page feature grid instead of on where their list sits on the cost-versus-deliverability curve. This article walks through the three approaches the way you should actually evaluate them: by what each one does with your sending reputation, your subscriber data, and your monthly bill.

The three architectures, defined by where sending happens

Strip away the marketing and every WordPress newsletter setup is one of these:

Self-hosted, BYO transactional relay. A plugin (MailPoet, FluentCRM, Newsletter, The Newsletter Plugin) stores subscribers in your WordPress database and composes campaigns in wp-admin, but hands the actual SMTP/API delivery to a transactional provider — Amazon SES, Postmark, SendGrid, Brevo, or Mailgun. You own the list, the relay owns the pipes.
Full external ESP. Kit (formerly ConvertKit), MailerLite, Mailchimp, Beehiiv, or ActiveCampaign owns the list, the editor, the automations, the sending IPs, and the reporting. WordPress just feeds it signups through an embed or a sync plugin.
Hybrid / headless capture. WordPress runs the public-facing capture and preference layer (forms, double opt-in, a logged-in member's email settings) while the list of record and the sending live in an external system, kept in sync via API or a plugin connector.

The dividing line that matters is not "plugin vs. SaaS." It is whether your domain's reputation is being built on a shared pool you don't control, a dedicated setup you must warm yourself, or an ESP's managed infrastructure.

Why deliverability is the real decision, not the editor

People shop newsletter tools by looking at the drag-and-drop editor and the automation canvas. Those are real, but they are the part you'll stop caring about after week two. The thing you'll care about for years is whether your mail lands in the inbox, and that is almost entirely a function of authentication and engagement, not which tool drew the email.

As of 2026, the Gmail and Yahoo bulk-sender requirements that took effect in early 2024 are simply table stakes. If you send to more than 5,000 recipients a day to those domains you must have:

SPF and DKIM both passing and aligned with your From domain.
DMARC published — even a p=none record satisfies the policy requirement, though p=quarantine is the real goal.
One-click List-Unsubscribe (RFC 8058) in the headers, not just a footer link.
A spam-complaint rate held under 0.3%, ideally below 0.1%, measured in Google Postmaster Tools.

This is where the architectures diverge sharply. A full external ESP handles authentication for you (you add a few DNS records they generate) and sends from IPs whose reputation they manage across thousands of senders. A self-hosted-plus-SES setup makes you responsible for warming the IP or sending pool, monitoring bounces, processing complaint feedback loops, and suppressing bad addresses. Amazon SES will happily let you wreck your own reputation in a weekend if you import a stale list and blast it.

The economics, with current numbers

Here is the trade-off in plain figures, because it's the whole game above a few thousand subscribers.

Transactional relays bill by volume sent. Amazon SES is roughly $0.10 per 1,000 emails. Postmark sits around $15/month for 10,000 sends with excellent deliverability and logs. Brevo and Mailgun have free or low tiers that scale on volume. So a 20,000-subscriber list sending four campaigns a month — 80,000 emails — costs about $8 on SES, plus the one-time and ongoing effort of running it.

External ESPs bill by subscriber count, which is brutal as you grow because you pay for inactive subscribers every single month whether you email them or not. Typical 2026 list-size pricing: ~$25-35/month at 5,000 contacts, ~$50-90 at 10,000, ~$140-250 at 25,000, and four figures past 100,000. Kit and MailerLite tend to undercut Mailchimp; Beehiiv prices by a different model and bundles monetization. The number you pay buys you managed deliverability and zero ops time.

So the cost ratio between self-hosted and ESP at 25,000 subscribers is easily 20:1 or worse. That gap is exactly why hybrid setups exist — they're the attempt to capture SES-level economics while borrowing ESP-grade deliverability discipline.

Choosing by where your list actually is

Under ~2,000 subscribers

Cost is noise here. The free tiers of MailerLite (up to 1,000) or Kit (up to 10,000 on its free plan, with paid features gated) cover you, and the deliverability is handled. Pick a full external ESP and spend your attention on writing something worth subscribing to. Running your own SES sending at this scale is a hobby tax with no payoff.

~2,000 to ~25,000 subscribers

This is the genuinely contested band. If newsletter revenue or engagement is core to the business and you have no appetite for DNS and bounce-handling, stay on an ESP — the $50-150/month is cheap insurance. If you're cost-sensitive and reasonably technical, FluentCRM or MailPoet paired with Amazon SES is the sweet spot: your data stays in WordPress, integration with WooCommerce or LearnDash is native, and you'll pay single-digit dollars in sending. The catch is you must own deliverability: warm the volume gradually, watch Postmaster Tools, and prune non-openers.

Above ~25,000 subscribers

The ESP bill is now a real line item and self-hosted economics are compelling — but only if you treat deliverability as an operational discipline. This is hybrid territory: WordPress (FluentCRM, or a custom capture layer) for forms, preferences, and segmentation, with a serious relay behind it. Consider a dedicated IP (available on SES, SendGrid, Postmark) once you're consistently sending enough volume to keep it warm — a dedicated IP that sends sporadically performs worse than a shared pool, so don't reach for one too early.

The trap nobody warns you about: migration resets your reputation

Whatever you choose, the most expensive day is the day you switch. Exporting subscribers is trivial — email, name, custom fields, tags all move cleanly. What does not move is engagement history and the receiving providers' learned sense that your mail is wanted. When you import 40,000 addresses into a new sending system, that system has zero reputation for your sending pattern, and the first few campaigns can sail straight into spam while it earns trust.

Mitigate it the way professional senders do: don't blast the full list on day one. Warm up by sending first to your most engaged segment (opened in the last 30 days), then expanding over one to two weeks. Suppress anyone who hasn't engaged in six months before you migrate — a smaller, hotter list outperforms a big, cold one on every metric that controls inbox placement. Set DMARC up before the cutover, not after.

A practical default

If you want one recommendation to anchor against: start on a managed ESP, and only move to self-hosted FluentCRM-plus-SES when your ESP invoice crosses roughly $100/month and you have someone willing to own deliverability. Below that line, the time you'd spend babysitting SMTP is worth more than the money you'd save. Above it, with the right discipline, the savings compound for years — and your subscriber list lives in a database you fully control rather than rented inside someone else's platform.

WordPress Newsletters: Plugins, ESPs, and Hybrid Approaches