How to Grow a Newsletter: Recommendations, Swaps, Paid Ads, Referral and What Each Subscriber Costs

To grow a newsletter, start with recommendations, cross-promotion and paid acquisition. Then add owned content, search and referral, which compound more slowly. Free swaps cost neither side anything. Paid recommendations typically run $2 to $7 per confirmed referral, and one practitioner targets a cost per acquisition (CPA) of $1 to $2 on Meta ads. Judge every channel on 30-day opens and retention, not on signup cost.

How do you grow a newsletter from zero

A newsletter grows through four channels: recommendations and cross-promotion, paid acquisition, owned content and search, and referral. Start with the first two. They have the clearest published costs, and you can test both inside a month.

Outside Substack, cross-promotion comes in a free form and a paid form. With free swaps, partner publications grow each other's subscriber lists at no cost to either side. With paid recommendations, the publisher that recommends you earns a commission that typically ranges from $2 to $7 per confirmed referral. On SparkLoop, a newsletter that wants to be recommended runs a Partner Program starting from $100 a month.

In our reading, Paid ads on Meta have become cheaper for operators who test their creative. One practitioner at Inbox Collective wrote in October 2025 that she once celebrated acquiring an email address for $1.50 and now regularly gets subscribers for under a dollar. She also warned that extremely low CPAs often indicated the wrong audience. CPA means cost per acquisition. The range she aims for is a CPA between $1-$2.

Owned content and referral compound more slowly. Search-friendly archives and a referral loop pay off once you have readers who share. Measure referral with a k-factor before you spend on prizes.

Channel	What the cost looks like	First step this week (our reading)
Free swaps	No cost to either side	Pitch three newsletters of similar size
Paid recommendations	$2 to $7 per confirmed referral	Set a small monthly test budget
Meta ads	Target CPA of $1 to $2	Run two plain text ads against each other
Content and referral	Your time	Add a referral link to the welcome email

A worked example

These are example inputs, not benchmarks. A $500 test budget at a $2 CPA buys 250 subscribers. The same $500 spent on paid recommendations at $5 per referral buys 100 subscribers. The ad route looks cheaper on day one. In our reading, the cost only means something once you compare 30-day open rates for each cohort, because a cheap subscriber who never opens adds nothing to your payback period.

How newsletter recommendation networks work

Each platform runs the same basic move. A reader subscribes to you, and at that moment they see a short list of other newsletters you vouch for. One click adds them to those lists too. The differences are where the picks appear, how many you can show and which plan you need.

Platform	Where readers see your picks	Plan or limit
beehiiv	Right after signup	Free on all plans, paid on Lite and Pro
Kit Creator Network	When people subscribe to your list	Up to 20 creators at once
Substack	Subscribe flow, homepages and automated emails	Built into the platform

What you control on each

On beehiiv you can pin a newsletter, rotate between several, or let smart matching fill a slot. On the paid side, you are only charged when a subscriber passes verification, and you can set a minimum engagement rate so that spending pauses if a partner falls below it.

Kit tags everyone who arrives this way with a form called "Creator Network." That form lets you put them in their own segment and send them a dedicated welcome sequence. One catch: if you already use SparkLoop's Signup Flow, Kit's Recommendations screen will replace your Signup Flow widget on any form where both are active.

In our reading, recommendations mostly fill the free list, and the paid conversion still has to come from your own content.

Fewer picks send more readers to each

With 100 new subscribers daily and your recommendation conversion rate is 20%, you send 20 readers out. Spread across 10 creators, each gets about two a day. Recommend only two, and each one will receive an average of 10 subscribers per day.

Run this calculation before you pitch partners. Here is an example, not a benchmark: at 500 daily signups and a 20% recommendation conversion rate, you send out 100 readers a day. Split across 5 partners, that is 20 readers each per day out of your 100. That daily number is your offer when you ask a partner to recommend you back.

We think a short list of partners whose readers overlap with yours works better than filling every available slot. Track what each partner sends back over a 30-day cohort, and swap out any partner whose referred readers stop opening.

What paid recommendations and boosts cost per subscriber

Paid recommendations turn the free swap from the previous section into a priced transaction. You set what one subscriber is worth to you, and other newsletters show your publication to their new signups in exchange for that fee.

beehiiv: a price per verified subscriber

On beehiiv, paid recommendations were previously called Boosts. You set a cost per acquisition (CPA), which beehiiv labels cost per subscriber. A 20% fee covers operational and Stripe transaction costs and is already factored into the marketplace price. beehiiv's own example: you pay $2.50 per subscriber acquisition, and the recommending publication is paid $2.

Setup has a few gates:

• Only a workspace Admin or Owner has access to paid recommendations, so a first marketing hire may need the founder to grant the role.
• You must complete Stripe Identity Verification before creating an offer.
• $20 is the minimum that you are able to deposit.
• Each publication can have only one live paid offer at a time.
• If your balance gets too low, your offer can pause, and a paused offer doesn’t appear in the Marketplace.

SparkLoop: a monthly budget

On SparkLoop, growing through paid recommendations means running a Partner Program. You set a monthly budget and pay only for confirmed, engaged subscribers. Beyond the entry Optimized plan covered earlier, the Custom Partner Program starts from $2,000/month.

	beehiiv paid recommendations	SparkLoop Partner Program
Pricing unit	Your CPA per verified subscriber	Monthly budget, paid per confirmed subscriber
Fee	20% built into the price	20% withheld from the earning side
Entry amount	$20 minimum deposit	Custom plan from $2,000/month

A worked example from both sides

These are example inputs, not benchmarks.

As the buyer: on beehiiv, a $5 CPA with the 20% fee means recommending publications see $4. A $500 budget at $5 per verified subscriber buys 100 subscribers. Check that figure against your own payback period before raising the CPA.

As the earner: the commission range cited earlier is what a recommending newsletter collects. SparkLoop says you'll make an extra $2-3 per new subscriber on average. At $2 each, 1000 new signups in a month would gross $2,000. SparkLoop withholds a 20% commission plus small processing fee on your monthly payouts, so $2,000 gross becomes $1,600 before processing fees.

In our reading, the earner's side is worth running even if you never buy. A small newsletter can use recommendation income to fund its own CPA on the buying side.

How to run a cross-promotion swap that brings in readers

A swap is two newsletters promoting each other to their own readers. It costs no cash, which is why it gets run carelessly. The method below has four steps, and each one guards against a specific way swaps fail.

Step 1: Pick a short list of adjacent partners

Do not email every newsletter in a directory. Growth In Reverse puts it plainly: "Instead of finding 100 people to partner with, find 3-5 who are perfect for your audience and optimize those."

Adjacent beats identical. According to growtoro, swaps convert "on topical adjacency, not vague demographic overlap." A founder newsletter pairs with one about hiring or fundraising, not with another general startup letter. The same source suggests partners within roughly 0.5x to 2x your list size, so the trade stays fair.

Step 2: Write their blurb for them, as an ad

Growtoro treats the blurb as "a 40-word ad" and says to hand it over ready to paste. A usable structure:

• Who it is for: name the exact reader.
• What they get: one concrete deliverable and its cadence.
• Why trust it: a one-line personal endorsement from the partner.

Growth In Reverse also writes several lengths, so the copy fits a sponsor slot or a one-line mention.

Step 3: Give every swap its own tracked link

Then log each swap on its own row.

Metric per swap	What it tells you
Click-through rate on the blurb	Whether the copy works
Signups divided by partner send size	Audience fit
30-day retention against your baseline	Audience quality

The retention row matters most. Growtoro's example: a swap with 400 signups that churn in three issues is worse than one with 120 who stay. Track it with the same retention view you use for any other channel.

Step 4: Do the math before you say yes

Growtoro reports that a well-matched swap converts 2% to 6% of the partner's audience, and a bad one converts under 0.5%. Take a partner sending to 40,000 readers, the list size growtoro uses in its own example.

• At 2% of 40,000, the swap brings 800 subscribers.
• At 6% of 40,000, it brings 2,400.
• At 0.5% of 40,000, it brings 200, so a poorly matched swap lands below that.

That spread, in our reading, is why partner selection deserves more of your time than the send itself. Run your first swap with a new partner as a test, judge it on the conversion and retention rows, and repeat only with partners who clear both. Treat it like any other growth experiment: one variable, a written prediction, a decision rule.

Paid social ads can bring newsletter subscribers in under a dollar

Paid social is the channel where money buys signups directly, and Meta is where most newsletter operators start. beehiiv's growth guide lists sponsored newsletter ads and paid social media ads among its quick wins, next to the paid recommendations covered earlier. It also suggests testing paid recommendations, sponsored placements, and targeted social advertising once a list has early traction.

The problem is that most Meta advice was written for stores. Inbox Collective ran campaigns for multiple newsletter-first media businesses over nine months and found that e-commerce strategies did not carry over directly to newsletter sign-ups. Its cost path is the useful benchmark. An email address at $1.50 had once been a win. In its 2025 write-up, it reported that it regularly brought in subscribers for under a dollar.

What the published numbers show

Ad format	Reported cost per subscriber	What it does
Contrast ad	$0.65 to $1.60	Shows what is specific to your city or niche
Testimonial ad	$0.88 up to $2	Lets a real subscriber quote do the selling

Rough beat polished. Quick text-on-background ads produced more subscribers at a lower cost than designed ones.

Cheap is not automatically good. Inbox Collective found that extremely low CPAs, meaning cost per acquisition (CPA), often signaled the wrong audience. It put its target range at a CPA between $1 and $2. Price follows competition, because Meta auctions its ad slots. In our reading, sub-dollar results come from narrow or local audiences that few advertisers bid on, not from cutting quality.

Timing matters too. Avoid major holidays and the days after them, when CPAs tend to spike.

A first test for this week

1. Write two plain ads: one contrast ad naming your niche or town, one testimonial using a reader's own words.
2. Set a ceiling. As an example, not a benchmark, a $500 budget at a $2 CPA buys 250 signups.
3. Tag every ad signup so it forms its own cohort.
4. Judge the cohort on retention and opens after 30 days, not on day-one CPA. A cheap cohort that churns fast cost more than it looked.
5. Treat it as a growth experiment. Keep the winning format, then test day and time against it.

Newsletter cost per subscriber by channel

Earlier sections priced each channel on its own. This table puts them side by side so you can copy it into a planning doc and fill in your own numbers as tests come back. The published cost column carries only what the sources state. The effort, timing and owner columns are our reading.

Channel	Published cost per subscriber	Effort (our reading)	Time to first signal (our reading)	Who does it (our reading)	What to measure
Free recommendations	No cost to either side	Low: set up once	First week of signups	Founder	Signups per partner, 30-day retention
Paid recommendations	Paid per confirmed subscriber; budgets from $100/month on SparkLoop's Optimized plan	Low to medium: set budget, review partners	First billing cycle	Founder or first marketing hire	Cost per confirmed subscriber, retention by source
Swaps	No cash; you give a placement in return	Medium: outreach, blurbs, tracking	After the partner's send	Founder	Signups divided by partner send size, retention
Meta ads	As low as $0.70 in some cases; see the target range above	Medium to high: creative and timing tests	Days	Growth marketer	Cost per acquisition (CPA), retention, unsubscribes
Owned content	No published figure	High: writing and distribution	Slow	Founder or editor	Signups per post or page
Referral	Set by your reward cost	Medium	After the first sends with the link	Founder	Referrals per active reader

Referral has its own mechanics, rewards and math. The referral program playbook covers it.

How to fill the table this week

1. Tag every source. SparkLoop lets you create a unique tracked link for each email you send. Use one link or form per swap partner and per ad set too.
2. Start swaps small. A shoutout swap is lowest effort, lowest conversion, which makes it a cheap way to vet a partner. One playbook suggests 1 to 2 one-off dedicated swaps per month for new partners.
3. Add auto-recommendations once a partner proves out. That format catches readers at peak intent, right after they subscribe.
4. Schedule ads by audience. A business newsletter might perform best on Monday mornings.

Worked example

These are example inputs, not benchmarks. Suppose $500 goes to ads at $2 per subscriber. That buys 250 subscribers. Now suppose a swap partner sends to 10000 readers and you get 100 signups. That is a 1% conversion on send size, and it costs no cash.

Raw cost does not settle the comparison, in our reading. Put each channel's 30-day retention next to its cost, then move budget toward the channel whose retained subscribers cost least. The growth model template has room for both columns.

How to measure subscriber quality by source and when paid growth stops paying

List size tells you how many people signed up. It does not tell you how many read. Morning Brew's growth team treated an individual subscriber's open rate, especially in the first two weeks, as its most important newsletter metric. Their reasoning: "We found that that's typically very indicative of how engaged they'll be over their full lifetime."

The gap between channels is large. At Morning Brew, open, click and referral rates varied as much as two or three times across channels,

Tag every subscriber by source

Give each channel its own tag, form or landing page, so every new subscriber carries a source label from day one. Some platforms do part of this for you. Kit associates every subscriber received through Recommendations with a Form called "Creator Network." On beehiiv, auto-clean was replaced by subscriber verification tags, and for paid recommendations you are only charged for subscribers who pass verification.

Then group each source into a two-week cohort and track the same three signals for each one.

Metric per source cohort	What it tells you (our reading)
Opens in the first two weeks	Whether the channel brings readers or just addresses
Clicks	Whether readers act on what you send
Unsubscribes	Whether the promise that drew them matched the newsletter

For the longer view, read each cohort against a retention curve, as explained in how to read retention curves.

Worked example: cost per engaged subscriber

This is an example, not a benchmark. Channel A costs $2000 and brings 1000 subscribers, so each costs $2. Channel B costs $2000 and brings 2000 subscribers, so each costs $1.

Now count who opened in the first two weeks. In Channel A, 500 of 1000 opened, so its cost per engaged subscriber is $4 ($2000 divided by 500). In Channel B, 100 of 2000 opened, so its cost per engaged subscriber is $20 ($2000 divided by 100).

The cheaper channel cost five times as much per reader. Morning Brew's growth lead made the same point: "Sometimes it makes more sense to invest more in the slightly more expensive channel."

Our rule for when paid growth stops paying

Divide each channel's spend by its engaged subscribers, never by total signups. Then compare that figure with what one engaged subscriber returns, through sponsorship, paid upgrades or referrals, inside your payback period.

Once cost per engaged subscriber exceeds that return, cut or pause the channel. In the example above, if an engaged reader returns $5 within your window, Channel A at $4 keeps its budget and Channel B at $20 loses it. Rerun the check every month, because a channel's cost per engaged subscriber drifts as you spend more on it.

A 90-day plan for a startup newsletter starting at zero

This plan maps 90 days onto the stage thresholds in beehiiv's growth guide, which splits the early climb into 0 to 100, 100 to 1,000 and 1,000 to 5,000 subscribers. For the first stage, that guide says "clarity and direct outreach matter more than complicated growth systems". So the first month is manual on purpose.

Days	What to do	Who does it	What it produces	Signal it worked (our reading)
1 to 30	Email personal contacts one by one, publish a landing page, turn on free recommendations	Founder	First readers and a live signup page	Signups arrive from people you never emailed
31 to 60	Run two or three swaps with adjacent newsletters, launch a referral program	First marketing hire	Subscribers tagged by partner, first referrals	One partner worth a repeat swap
61 to 90	Run a small paid test with a tracked link per source	Growth marketer	Cost per subscriber by source	The paid cohort opens close to the organic cohort

Days 31 to 60: swaps and referral

Keep the partner list short. Growth In Reverse advises you to "find 3-5 who are perfect for your audience" instead of cold-emailing dozens. Give each partner a link with its own parameter, such as ?utm_source=theirnewsletter&utm_medium=email, and a ready blurb written in a long and a one-line version.

Set expectations before you start. One writer reported 465 subscribers from running cross-promotions in the first eight months of his newsletter. In our reading, Swaps add readers steadily. They do not double a list in a month.

Launch the referral program once the welcome email is final, so new readers see the share link in the first message they open.

Days 61 to 90: the paid test

Cap the budget before you buy anything. As a worked example, not a benchmark: Rowan Cheung paid $2.00 per subscriber through paid recommendations, so at that price a $500 test budget buys 250 verified subscribers.

On beehiiv, paid recommendations are available on Lite and Pro plans. The network also lets you choose your minimum engagement rate, and spending on an agreement pauses automatically when it falls below that rate. Set it on day one of the test, not after the money is gone.

At day 90, compare each source as a cohort using the two-week open check from the measurement section. Then keep only the channels that pass. Log every step above as a test in the growth experiment backlog, with a start date, an owner and the signal you will judge it by, following the method in how to run a growth experiment.

What this page does not cover

This page does not cover referral program mechanics and reward design in detail, which the referral program playbook handles, and it does not compare or choose between newsletter platforms. It also leaves out how to price sponsorships or paid upgrades.

Related on Race to Scale: Founder-led marketing on LinkedIn and X: the cadence, formats and what to measure; Referral program examples that worked, and the mechanics behind them; Waitlist launch strategy: how the referral queue and the gated waitlist work, what converts, and when a waitlist backf.