Newsletter Growth skill for Codex

A newsletter is run well when one sentence explains who it is for and what they get each issue, every subscriber chose to be there (and confirmed it), new readers get a welcome that sets expectations, and growth is measured as net active readers, not total sends.

Diagnose before growing

Ask for or export:

• Total subscribers, new per week, unsubscribes and bounces per week (net growth).
• Clicks per issue and unique clickers; 30/90-day active readers if the platform reports it.
• Acquisition sources (platform source reports, UTM on signup pages, referral counts).
• Current promise: the subtitle or tagline on the signup page.

Symptom	Likely problem	First fix
Signups fine, clicks falling	Content drifting from the promise, or cadence too high	Re-read the last 10 issues against the promise; survey readers
High unsubscribes after issue 1-2	Welcome missing, or signup promise oversold	Write a welcome sequence; match signup copy to actual issues
Few signups from existing traffic	Weak or hidden forms	Signup above the fold on posts, end-of-post form, dedicated landing page
Growth only from one source	Fragile distribution	Add referral and recommendation channels
Spam folder placement	Poor engagement, no authentication, bought or bot signups	Double opt-in, sunset inactives, check SPF/DKIM/DMARC (email-marketing)

Positioning and the signup page

• One-sentence promise: <What> for <who>, every <cadence>, so they can <outcome>. “A 5-minute Friday brief on Postgres performance for backend engineers.”
• Signup page: promise as headline, 3 bullet examples of recent issues, social proof (subscriber count only if impressive and true; named readers with permission), sample issue link, one email field.
• Embed forms where readers already are: end of every blog post, a slim inline form mid-article, the site header or footer, and a dedicated /newsletter page for links.
• Find existing forms in the repo before adding new ones: rg -n "beehiiv|substack|kit.com|convertkit|ghost|buttondown|newsletter" src/ app/ components/.

Signup quality

• Double opt-in by default; it keeps bot signups and typos out and protects deliverability.
• Bot “list bombing” attacks submit thousands of real victims’ addresses: enable the platform’s bot protection, rate limit custom forms server-side, and keep double opt-in on.
• Lead magnets (template, checklist, mini-course) raise signups but attract freeloaders; track the 30-day click rate of magnet signups separately and keep the magnet on-topic.
• Never import scraped or purchased lists, and never auto-subscribe people who only bought a product or created an account without telling them clearly at signup.

Welcome sequence

Default: 2-4 emails over the first 7-10 days, then into regular issues.

Email	When	Content
1	Immediately after confirm	Thanks, what to expect (cadence, topics), best past issues, ask them to reply with one question (replies help inbox placement)
2	Day 2-3	The single most useful piece you have written; how to whitelist / move to Primary
3	Day 5-7	Who writes it and why; invite to the referral program or community
4 (optional)	Day 9-10	Paid tier or product offer, if any, with a clear free alternative

Growth channels (rank by fit, then work the top two)

Channel	Mechanics	Notes
Content and SEO on the site	Articles that rank, each with an inline form	Slow, compounding; pair with seo-content-writing
Referral program	Reward readers for confirmed referrals (beehiiv built-in, SparkLoop, or custom)	Reward tiers that cost little: guide, sticker, community access; count only confirmed and active referrals
Recommendations	Substack Recommendations, beehiiv Recommendations, Kit Creator Network	Swap with newsletters of similar audience and size; check their open/click quality
Paid recommendations	beehiiv Boosts, SparkLoop Partner Network	Pay per confirmed subscriber; measure 30-day engagement of those cohorts before scaling
Cross-promotions and guest issues	Write for another newsletter in your niche	One of the best quality sources
Social	Post the best insight from each issue natively with a signup link	See social-media-content
Paid ads	Meta lead ads, X, Reddit	Only once you know subscriber value; expect lower engagement

Rule: judge every source by the 30-day engagement of the subscribers it brings, not by the count.

Issue format and cadence

• Consistent day and time; consistency beats frequency.
• A recognisable structure readers can scan: one main piece, then 3-5 curated links with a sentence on why each matters.
• Subject line: specific to this issue’s best item, not “Issue #47”.
• Each issue has one primary action (read, reply, share, click the main piece).
• Plain readable HTML; most reading happens on phones. Test in Gmail and Apple Mail.
• Include a referral or share prompt in every issue, near the end.

List hygiene

• Opens are inflated by Apple Mail Privacy Protection; define “active” by clicks (or the platform’s engagement model) within 90 days.
• Re-engagement: send inactive subscribers (no clicks in 90-180 days) a short “still want this?” email; remove those who do not respond within 2 weeks.
• Removing or archiving subscribers is hard to undo: export the segment first, show the user the count, and get confirmation before deleting.

Monetisation

Model	When	Watch
Sponsorships	Engaged audience in a niche advertisers want	Price on CPM of engaged readers or flat per placement; label ads clearly as sponsored
Paid tier	Readers rely on it for work or decisions	Keep a strong free tier; paid conversion of a few percent of free readers is typical
Own products / consulting	Newsletter supports a business	Do not let promotion crowd out the promise

Disclose sponsorships and affiliate links clearly (FTC Endorsement Guides in the US; similar rules elsewhere).

Metrics to report weekly

| Week | New | Unsubs | Bounced/cleaned | Net | Active (90d clickers) | Click rate | Top source |

Net growth and active readers are the headline numbers. Investigate any issue with an unsubscribe rate well above the trailing average.

Gotchas

• Growing via giveaways: huge spikes, near-zero engagement, deliverability damage.
• Recommendation swaps with newsletters of a different audience; subscribers churn fast.
• Paying for Boosts without tracking cohort engagement; cheap subscribers can be dead weight.
• Changing sender name or address often; readers stop recognising it and filters reset.
• Sending every issue to inactive subscribers forever; it drags inbox placement for everyone.

Sending and publishing

Sending an issue, launching a referral program or importing subscribers reaches real inboxes. Prepare drafts, segments and settings; send, schedule or import only after the user confirms the final version and the recipient segment.

Verification

• Signup flow tested end to end: form, confirmation email, welcome email 1, correct tags/segments.
• Each acquisition source has its own tag or UTM and shows 30-day engagement.
• Weekly table shows net growth and active readers; trend reviewed monthly.

Output template

## Newsletter growth plan: <name>
**Promise:** ...
**Diagnosis:** metrics + symptom table rows that apply
**Signup surfaces:** list with changes
**Welcome sequence:** 3-4 emails (subject + body draft)
**Channels (next 90 days):** top two with targets and cohort-quality checks
**Hygiene policy:** active definition, re-engagement, removal schedule
**Awaiting approval:** sends, imports, deletions