When Should a Free Newsletter Start Charging? Lenny's Paid-Subscription Decision
Short answer: Start charging when you have published through at least two normal content cycles, can name a recurring outcome readers value, and can sustain a paid promise for six months without relying on launch excitement. Keep a useful free layer for discovery, offer one simple paid tier, and judge the first 90 days by retained paid contribution rather than subscriber count. Do not wait for an arbitrary audience size.
The usual debate is whether to charge from the first issue or build a large free audience. Both can work. The decision depends less on list size than on whether you have proved a repeatable publishing system and a reason to pay that is stronger than goodwill.
A paywall creates a business and a deadline at the same moment. If your consistency exists only while writing feels novel, early revenue can turn a useful practice into a monthly obligation you resent.
Lenny Rachitsky proved the habit before adding payment
Lenny Rachitsky began Lenny's Newsletter in June 2019 and added paid subscriptions in April 2020. In his account for Substack, he explains that he wrote for nine months before introducing payment and that charging created continuing pressure to publish.
In a separate first-year breakdown, Rachitsky reported roughly 15,000 free subscribers, 500 paid subscribers and about $65,000 in annual revenue before fees at that point. He also said newsletters can launch directly as paid, but argued that arriving with an audience that has already received value makes the transition easier.
The case does not establish nine months as a rule. Rachitsky had prior expertise, a technology audience with relatively high earning power and distribution from earlier writing. His figures are self-reported and represent one publication.
The useful decision is that he tested two capabilities separately. Free publishing proved he could produce useful work consistently. Paid publishing then tested whether a subset would fund it.
My view is that most free newsletters wait too long to ask for money and too little time to prove delivery. Six dependable issues with strong reader behaviour are more valuable than 20,000 disengaged addresses. Charge when the promise is ready, not when the vanity number looks respectable.
Use the Reader-Funded Conversion Gate
The Reader-Funded Conversion Gate has five conditions. A newsletter is ready to test payment when all five have evidence.
| Gate | Evidence | Failure signal | |---|---|---| | Publishing capacity | Issues arrive on the promised schedule across normal work weeks | Bursts followed by silence | | Reader outcome | Readers use the work for a named decision or result | General praise without action | | Repeat attention | The same suitable readers return, reply, save or follow links | List growth with no recurring behaviour | | Paid distinction | Paid work has a clear additional depth, frequency or access | Charging for previously promised basics without explanation | | Sustainable economics | Retained revenue covers production and service at realistic founder pay | Launch cash hides an ongoing loss |
You do not need a precise conversion forecast to run the test. You do need a minimum viable paid cohort. If 30 readers at the proposed price cannot justify the extra work, either the price, cost or offer is wrong for this stage.
Define what remains free
Free content is not a failed version of the paid product. It has three jobs: demonstrate judgement, help readers who cannot pay and introduce the publication to future suitable buyers.
Paid content needs a different job. Useful distinctions include deeper decision analysis, a more frequent operating update, proprietary reporting produced by you or access to a specific professional community. “More content” is weak because readers are already overloaded.
Write the boundary in one sentence. For example: free readers receive one practical analysis each month; paid readers receive a weekly analysis with the underlying decision arithmetic and a monthly question session. Then cost each promise.
Do not put every strong issue behind the paywall. If new readers cannot see your best thinking, acquisition depends on claims instead of evidence. Equally, do not make the paid tier a donation unless patronage is the explicit proposition.
Price the work, not another newsletter
Competitor prices provide context, not an answer. Start from the reader's recurring value and your delivery cost. A newsletter that helps an owner avoid one £500 purchasing error has a different ceiling from a weekend entertainment digest.
Offer one monthly and one annual option at launch. Annual payment improves cash timing but creates a 12-month delivery obligation. Recognise it internally across the service period instead of treating all launch receipts as spendable profit.
Model cancellations, platform and payment charges, research, editing, contributors, community moderation and a realistic rate for your time. The price must support the version readers are buying, not a future scale you hope will make it work.
Tax, consumer subscription, cancellation and data rules vary by country and customer location. UK and international publishers should check the current treatment of digital subscriptions and VAT with a qualified local adviser. Platform handling does not automatically remove every obligation.
Worked example: OpsLetter adds a paid tier
OpsLetter is a hypothetical UK newsletter helping small manufacturers manage purchasing and production. It has 6,000 free subscribers after 14 months and has published every fortnight. The founder plans a weekly paid operations teardown at £12 a month or £120 a year.
At launch, 180 readers choose monthly payment and 90 choose annual payment. The annual buyers provide £10,800 of cash, but OpsLetter treats £900 as the monthly revenue equivalent.
| Monthly economics | Arithmetic | Amount | |---|---:|---:| | Monthly-plan revenue | 180 x £12 | £2,160 | | Annual-plan monthly equivalent | 90 x £120 / 12 | £900 | | Total monthly revenue equivalent | £2,160 + £900 | £3,060 | | Platform and payment cost at 8% | £3,060 x 8% | £244.80 | | Research and editing | Given | £900 | | Founder production, 48 hours at £25 | 48 x £25 | £1,200 | | Support and administration, 10 hours at £25 | 10 x £25 | £250 | | Monthly contribution before shared overhead and tax | £3,060 - £244.80 - £900 - £1,200 - £250 | £465.20 |
The 270 paid readers represent 4.5% of the free list. That rate is not a general benchmark. It is simply this test's result.
Now apply three-month retention. If 20% of monthly subscribers cancel and no one replaces them, monthly-plan revenue falls to £1,728. With the other lines unchanged, contribution falls to £33.20. If producing weekly issues also takes 65 founder hours rather than 48, the offer loses £391.80 a month.
OpsLetter should not respond by chasing a bigger free list. It should learn why paying readers leave, reduce the production promise or raise a price whose value it can defend. The bottleneck is paid retention and workload.
Run the launch as a 90-day pricing test
Invite payment from the whole suitable list without pretending spaces are scarce. State the normal price, what paid readers receive, the first delivery date and how cancellation works. A genuine founding price can reward early uncertainty, but give it a clear duration and do not manufacture a countdown.
Track the cohort by plan:
| Measure | Why it matters | Decision use | |---|---|---| | Paid conversion | Strength of initial proposition | Diagnose price and promise | | Paid issue engagement | Whether buyers use the work | Improve topics and delivery | | Renewal or cancellation | Continuing value | Decide whether to scale | | Hours per issue | Capacity requirement | Reset frequency or scope | | Contribution per paid reader | Economic viability | Fund quality and acquisition | | Free-list growth | Discovery remains healthy | Protect the top of the funnel |
Email opens are increasingly imperfect because privacy features can record activity that did not occur. Combine them with replies, clicks, attendance, renewals and voluntary feedback. Do not punish quiet readers who continue to pay and use the work offline.
Do not let paid readers dictate every issue
Payment deserves service, not editorial control. A few vocal subscribers can pull the publication towards bespoke advice that most readers do not need.
Collect questions, then choose topics where the same decision appears across the intended segment. If one company needs a confidential analysis, that is a different product with different pricing and capacity.
Protect the publication's independence. Sponsorship, affiliates and paid subscriptions create different incentives. Disclose commercial relationships and keep the editorial promise clear.
Make the decision in 16 weeks
Weeks one to four: define the free and paid jobs, cost six months of delivery and interview active readers about decisions they make. Weeks five and six: publish the offer, price and sample of paid depth.
Weeks seven to sixteen: deliver at the promised cadence and track the first cohort. Review cancellations individually without offering unplanned custom work to save every account.
At day 90 after launch, continue if paid contribution is positive, workload is sustainable and retained readers use the distinct value. Otherwise change one of price, promise, cadence or audience, then retest. Do not hide the result by counting annual cash as earned profit.
Related guides
Frequently asked questions
How many free subscribers do I need before charging?
There is no universal minimum. You need enough suitable, engaged readers to form a meaningful paid cohort at your required price. Work backwards from economics. If the paid tier needs 100 readers and you reasonably expect 2% conversion, you need access to roughly 5,000 comparable free readers. That 2% is an assumption to test, not a benchmark. A specialist publication may convert 80 of 800 readers, while a broad entertainment list may need far more reach. Use repeated reader behaviour and bottom-up arithmetic rather than a prestige threshold.
Should I put the whole newsletter behind a paywall?
Usually keep a strong free layer unless readers already know and trust your work through another channel. Free issues demonstrate quality and create discoverability; paid issues must provide a distinct recurring value. A complete paywall can work for scarce professional information or an established author, but it makes the first purchase depend on reputation and samples. Test the boundary explicitly. If free readers get the full paid outcome, conversion will be weak. If they get nothing useful, future acquisition and trust may weaken.
Is monthly or annual pricing better at launch?
Offer both when you can honour a year of delivery. Monthly lowers commitment and reveals churn sooner. Annual improves cash timing and may attract committed readers, but it creates a longer service and refund exposure. Compare plans using monthly revenue equivalent, not launch cash. Keep enough annual receipts reserved to produce future issues and meet lawful customer refunds. Subscription, tax and cancellation rules vary by customer location, so confirm current requirements with qualified local advisers before relying on annual prepayment.
How should I choose the first paid price?
Choose a price that fits the value of the reader's recurring decision and covers delivery at a plausible early cohort size. Interview readers about alternatives and consequences, then test one clear number rather than six confusing tiers. Check comparable publications only to understand expectations. Do not automatically underprice because the newsletter is new. A low price can attract weak-fit buyers and make research uneconomic. State when an introductory price ends, and assess conversion together with retention and contribution after at least three billing cycles.
What if readers say they love it but will not pay?
Believe the payment behaviour. Praise may mean the free edition is pleasant, not that the paid outcome is urgent. Ask what they expected behind the paywall, which current expense or information source it would replace and what decision it would improve. Then change one variable: audience, paid promise, evidence or price. Do not keep adding hours to win people who have no budget or recurring need. Preserve the free publication only if it serves a deliberate acquisition, reputation or mission goal you can afford.
When should I stop the paid tier?
Stop or redesign it when retained revenue cannot cover the promised work at realistic pay, cancellations reveal no recurring outcome, or the obligation damages the quality of the free publication and your main business. Honour existing annual commitments, communicate dates and provide refunds where required. Do not continue because closing feels embarrassing. A finite paid experiment can reveal that sponsorship, consulting, events or a free publication better suits the audience, although each alternative needs its own economics and capacity test.
Comments (0)