How to get your first subscribers for a newsletter

Publish four or five issues in public before you ask a single person to subscribe. Not drafts, not a landing page with a promise - actual issues, on the open web, with dates on them. Nobody can evaluate a newsletter that does not exist yet, and "sign up to find out what it will be like" is the weakest offer in publishing.

This is the one product type where doing the work and marketing the work are the same activity. Each issue is at once the product, a sample, a page that can rank, something a reader can forward, and the thing another writer needs in order to justify recommending you. Nothing else in this cluster has that property, and almost every tactic below depends on it.

Your peers are your best distribution, and they are reachable

A competing SaaS product will not promote you. A newsletter of similar size on an adjacent topic frequently will, because recommending you costs them almost nothing and the exchange is symmetric. Cross-promotion - a personal recommendation in your own voice, traded with someone at roughly your stage - is the fastest early channel here and costs nothing but a good-faith email. The awkward part is needing something to trade, which is another reason the archive comes first.

The platforms have built recommendation into the product

Most modern newsletter platforms ship some form of recommendation mechanism: one writer suggesting another at the moment of signup, when intent is at its peak. It compounds in a way a social following does not, because every new subscriber arrives already in the act of subscribing. If your platform has it, switching it on and curating who you recommend is among the best effort-to-effect ratios available anywhere on this site.

The archive is a search asset most writers throw away

Every issue is a page. Titled properly, dated and published at a stable URL, an issue answering a specific question can rank for that question and be cited when someone asks an assistant about it, for years. Titled "Issue #14" and left inside an email service, it does none of that. Fixing this is an afternoon of configuration plus a naming habit, and it is the most common piece of free growth newsletter writers leave on the floor.

Your metrics have been quietly broken

Open rates stopped being a clean measurement when major mail clients began pre-fetching images on the reader's behalf, so a share of your opens are machines. Clicks, replies and unsubscribes survived. On a small list the most informative number is not a rate at all: it is how many humans reply when you ask them something directly.

The business model decides the writing, so choose it early

Sponsorship needs scale, pushing you toward broad topics. Paid subscriptions need intensity, pushing you toward a narrow audience who would genuinely miss you. A newsletter selling something else - consulting, a product, a book - needs neither, and is measured on conversations started rather than list size. Three different newsletters. Deciding late means rewriting the first year.

Almost entirely from places you are already standing. This is unromantic and it is the honest answer.

• People who already know you. Asked individually, not announced. A personal message with a link to a specific issue you think they will like converts at a rate no broadcast approaches.
• Communities where you already have a history. Not communities you join for this purpose. The forum you have posted in for two years is a warm audience; the one you joined last Tuesday is a place you are about to get moderated out of.
• Other people's audiences, by writing for them. Guest posts, podcast appearances, a genuinely good comment on a big publication. The transfer rate is high because reading a long thing you wrote predicts wanting more of it.
• Search, via individual issues. Slow to start, permanent once it works. An issue answering a specific question keeps recruiting long after you have forgotten writing it.
• Forwarding. One idea per issue a reader would want to send to a colleague does more than any share button. The button was never the constraint.

Ranked for a newsletter in its first few months, before it has enough volume for referral mechanics or paid acquisition to mean anything.

• Do not import contacts who never asked. Uploading a conference list or your address book generates complaints, damages the sending reputation of your domain, and starts you with an audience that resents you.
• Do not buy subscribers or run giveaways. You end up with a large list that never opens, which corrupts every number you would use to make decisions and, on some platforms, costs you money per dead address.
• Do not launch with a landing page and no issues. "Subscribe to find out" asks for trust you have not earned. Four public issues answer the question instead.
• Do not put the archive behind the signup form. Gating a free newsletter's back catalogue removes your search surface, your samples and your credibility in one move.
• Do not obsess over open rate. It is partly a measurement of mail clients now. Watch clicks, replies and unsubscribes instead.
• Do not change cadence while you are figuring it out. Readers form the habit around the rhythm, and an unpredictable arrival is functionally a different product each time.

A newsletter is easy to measure badly, because the platform dashboard leads with the number that no longer works.

• Replies per issue. The single most informative signal on a small list. A reply is unambiguous evidence a human read it and cared enough to respond.
• Click rate on the one thing you most wanted them to click. Not aggregate clicks across every link - the specific one, because that is the closest thing you have to a measure of intent.
• Net growth by source, written down manually. Which swap, which guest post, which issue. Ten rows in a text file beat a dashboard here, because the dashboard cannot tell you which recommendation did the work.
• Unsubscribes per issue, read as feedback rather than loss. A spike on one issue tells you something specific about that issue. A steady low rate is healthy and means you are saying things.
• Deliverability basics. Send from your own domain with authentication configured properly. Nothing else on this list matters if the email lands in spam, and this is the failure that hides itself most effectively.
• Ignore total subscriber count as a headline and open rate as a health metric. Five hundred people who reply beat five thousand who do not.

Newsletters compound more reliably than almost anything else here, and they compound slowly - a difficult combination for anyone used to shipping software. The first fifty subscribers take an uncomfortable amount of personal asking, the next few hundred come mostly from other writers, and the archive does nothing for months and then works indefinitely. The failure mode is not that the tactics fail. It is quitting at issue nine, which is roughly where the writing gets good and the numbers still look like nothing.

The pillar covers the manual stage every product type shares, and the channel picker narrows which two of the above are worth your first month.

The channel picker asks who you are writing for and how they currently find things to read, then names the two or three channels worth your first month. For newsletters it mostly separates the ones that should be swapping with peers from the ones sitting on an archive that could rank and has never been published properly.

Or get the 30-day post-build playbook Send it