Growing a newsletter without becoming a growth-goblin

There is a version of newsletter growth that works and a version that works. The second version — pre-ticked consent boxes, exit-intent pop-ups that reappear daily, a "welcome gift" that is a coupon for something nobody wanted, bought lists with names like james7229@ — produces a big number that decays like fruit. The first version is slower to build and then compounding for a decade.

We've argued elsewhere that a newsletter should be designed as an engine — with a promise, an issue architecture, and loops baked in. This piece is about the part that determines whether the engine's fuel is real: acquisition without trickery, a welcome sequence that earns the second open, referral mechanics that don't bribe strangers, list hygiene as a growth lever, and measurement that survives the death of the open rate. Honest growth is not an ethical luxury. It is the only version with good unit economics.

The sign-up page: your real conversion problem

Most newsletters don't have a growth problem; they have a sign-up page problem. The modal pop-up everyone A/B-tests to death converts a fraction of a percent. The dedicated landing page — the one you link from socials, from guest articles, from the footer of every forwarded issue — often converts twenty times better, and nobody writes it.

The page is one screen with five jobs:

1. Restate the promise verbatim. The one-liner you settled when you designed the newsletter belongs in the headline, not "Subscribe to our newsletter," which is a category, not an offer.
2. Show the product. Three recent issues, readable full-text, behind one click — or better, one full issue rendered right there on the page. This single element is the biggest honest converter we know: people don't subscribe to newsletters, they subscribe to that issue they just read. It also quietly kills buyer's remorse unsubscribes, because nobody can claim the product was mis-sold.
3. State the cadence and the exit. "Weekly on Sundays. Unsubscribe in one click, always."
4. Prove other humans read it. A real testimonial or a subscriber count — only if both are true and durable. A count you'll be embarrassed to update next year is a liability, not an asset.
5. One field, one button. Email in, subscribe. No first name for fake personalisation, no checkboxes to untick. If you need segmentation, ask after the welcome sequence, when asking is a conversation rather than a toll.

When we rebuilt the sign-up flow for Hearthbrew's subscription club, moving the pitch off a footer module and onto a real page with a readable sample issue roughly tripled sign-ups per visitor — and, tellingly, raised ninety-day retention of new subscribers, because expectations were honest. Acquisition tricks optimise step one and tax step two forever. Showing the product optimises both.

Double opt-in belongs in this conversation, briefly: use it. Single opt-in grows the list maybe 20% faster and fills it with typos, bots and people who'll mark you as spam to make the emails stop. The "lost" subscribers were never yours; the deliverability damage is very much yours. Everything about deliverability as a growth lever starts with consent you can prove.

The welcome sequence: earning the second open

The first open is curiosity. The second open is a decision. The welcome sequence exists to win it, and we keep it short — three emails:

• Email 0 (immediately). Deliver whatever was promised, then set the table: what this newsletter is, the exact day it arrives, who writes it, and — critically — the ask to drag the email into the primary tab or reply with one word. That reply is not engagement theatre; it's a handshake with the spam filter, and it's proportionally more valuable than a click.
• Email 1 (day 2–3). The greatest hits: three issues that best represent the promise, each with a one-line argument for reading. This is onboarding, not content marketing.
• Email 2 (day 6–7). The human one: who we are, why this exists, and one genuine question. "What's the hardest part of X for you right now?" Replies here seed a year of editorial ideas and make the newsletter a correspondence rather than a broadcast.

Then the sequence ends and the regular cadence takes over. The sin to avoid is the twelve-email indoctrination funnel — welcome sequences are onboarding, not a hostage negotiation. This is onboarding design wearing an email costume, and it's governed by the same rules as lifecycle email architecture: each message has one job, and the programme degrades gracefully when someone ignores it.

Referral mechanics that don't bribe strangers

Referral programmes work when the reward matches the audience. They bribe strangers when it doesn't.

The failure mode is cash-equivalent rewards on a general-audience list: a $20 voucher per referral converts your readers into recruiters, and recruiters optimise for volume — posting their link in coupon forums, spamming colleagues, collecting sign-ups from people who will never open an issue and will drag your engagement rates into the spam folder. We've audited newsletter programmes where "growth via referral" was 40% dead addresses.

The mechanics that compound instead:

• Reward with identity, not cash. Recognition tiers (your name in the issue), access (a quarterly AMA, a private channel), or artifacts (the poster, the zine, the blend named after the top referrer). These attract people who love the thing, and those people refer people like themselves. Retention of referred subscribers follows the reward's meaning.
• Make the referral read like a recommendation. The pre-written share copy matters more than the reward. "I've been reading this for a year and it's the only newsletter I open on Sundays" refers better than "Join now and get 10% off!" — because the first is a favour from a friend and the second is an ad from a stranger.
• Threshold low, ceiling interesting. One referral should unlock something, so most subscribers can win. High tiers can be absurd on purpose — the five-referral mug nobody needs and everybody wants.

Track quality, not just count: ninety-day engagement of referred sign-ups versus organic. If referred subscribers engage worse, your mechanics are attracting bounty hunters, and the honest move is to change the reward, not to celebrate the number.

List hygiene: subtraction as a growth strategy

Here is the contrarian arithmetic nobody puts in the board deck: removing subscribers grows a newsletter. A list where 30% of addresses never open trains inbox providers to file you under "ignore," which quietly throttles delivery to the 70% who love you. Dead addresses are not neutral ballast; they are an active tax on the living.

The hygiene discipline, quarterly:

• Define "cold" honestly. No click or open (bearing in mind opens are now noisy — more below) in 90–120 days.
• Run one sunset sequence. Two emails: "Still want this?" with an easy re-confirm click. No guilt-trip subject lines, no "we'll miss you" theatre.
• Then actually unsubscribe the non-responders. The step everyone chickens out of is the step that matters. Lists we've cleaned this way typically shed 15–25% of addresses — and watch engagement rates, inbox placement and revenue per send all rise within a month. The total-audience number the marketing lead reports shrinks. Have that conversation before the sunset sequence, in writing, with the arithmetic attached.

Related hygiene: suppress chronic complainers and hard bounces instantly, keep role addresses (info@, sales@) off marketing sends, and never, ever buy or "partner-acquire" a list. A bought list is a deliverability incident with a receipt. If the delivery plumbing interests you, the full treatment is in our deliverability fundamentals piece.

Measurement after the open rate died

Apple's Mail Privacy Protection pre-loads tracking pixels, which means a large share of your "opens" — often 40–60% of a consumer list — are machines. Open rate is not "less reliable" now; it's a different metric wearing the old metric's name. This broke a decade of habits and, frankly, improved the industry's eyesight. What to watch instead:

• Click rate and, better, clicks per clicker. Was anything worth acting on, and for the people who act, how much?
• Reply rate. The most honest engagement signal email has, and the one spam filters reward most. This is why the welcome sequence asks for a reply and the issue architecture ends with a question.
• Conversion to the actual job. For most newsletters that's a downstream behaviour — trial starts, store visits, donations, repeat purchases. UTM everything, and reconcile it in the warehouse rather than trusting the ESP's cheerful attribution screen; our honest take on attribution applies to email doubly.
• Unsubscribe rate per send. The last un-gamed metric. Subscribers can't fake leaving. A stable, low unsubscribe rate with rising clicks is a healthy newsletter; a rising unsubscribe rate is the market editing your promise, and you should listen.

One warning against overcorrection: don't kill the trends along with the metric. Open rate is useless for counting but still useful as a ratio between your own issues — if Tuesday's "open rate" halves relative to your own baseline, something real happened, even if the absolute number is fiction.

Key takeaways

• The dedicated sign-up page is your real conversion lever — promise verbatim, a full readable sample issue, one field, and a stated cadence with a one-click exit.
• Double opt-in costs you ghosts and buys you provable consent. Take the trade every time.
• A three-email welcome sequence — deliver, greatest hits, human question — earns the second open; twelve-email indoctrination funnels earn resentment.
• Referral rewards should carry identity, not cash: bounties attract bounty hunters, artifacts attract evangelists. Measure referred-subscriber quality, not just count.
• Prune cold subscribers quarterly after a two-email sunset. Subtraction improves deliverability, engagement and revenue per send.
• After Mail Privacy Protection, measure clicks, replies, downstream conversion and unsubscribes — treat open rate as a relative trend signal, never a count.

FAQ

Won't double opt-in and aggressive hygiene slow our growth targets? They slow the vanity number and accelerate everything that pays. If a target is set in raw subscriber count, change the target — subscribed-and-engaged at 90 days is a metric your CFO and the inbox providers can both respect.

Should we gate content to grow the list? Sparingly, and only where the gated thing is genuinely a document — a calculator's PDF export, a research report. Gating ordinary articles trades long-term search compounding for emails of people who wanted the article, not you. We'd rather earn the subscription with a readable sample than extort it.

Pop-ups: always evil? No — intent-matched, once-per-visitor-per-month pop-ups on high-intent pages do fine. The sin is frequency and deception: reappearing daily, fake close buttons, "No thanks, I hate saving money" confirm-shaming. The test is simple: would you show this interaction to the person, on a stage, in front of your audience?

How big should a list be before a referral programme? Big enough that the programme's overhead is worth it — roughly 2,000 engaged subscribers — and small enough that you can still iterate the reward fast. Before that, just put a forward-link and a "share this issue" ask in the template and watch what happens.

What's a healthy unsubscribe rate? Under ~0.3% per send for a mature list sent weekly is comfortable territory; welcome sequences legitimately run higher because they're filtering. Worry about the trend and the reasons (ask, optionally, on the unsubscribe confirmation page), not a universal benchmark.