Is Your Free Tier Generous Enough? A Recommendation-First Test
A free tier can look healthy in a dashboard and still fail the only test that matters for organic growth: whether someone who never pays would send a friend to the app. This guide explains that test, why conversion rate alone cannot answer it, and how to find the free-tier boundary with two metrics tracked side by side.
Why conversion rate cannot answer the question
Freemium is meant to start a word-of-mouth engine: give away the core product, let users build a habit, and they bring in others. Many apps never get there. The free tier behaves like a stripped-down trial, people leave before a habit forms, and no one recommends anything.
The usual diagnostics miss this because they describe payers. Conversion rate, paid penetration and LTV all measure what paying users do. None of them tells you whether your non-paying users are an asset or dead weight.
Opal CEO Kenneth Schlenker, speaking on the Sub Club podcast, offered a different question: would a non-paying user recommend the app to a friend? If the answer is no, the free tier needs to give away more. As he put it, the freemium dynamic only pays out when free users are recommending the app; otherwise it does not work.
The mechanism is simple. Freemium works only if free users stay long enough to build a habit and then recommend the app. A free tier that is too restrictive — missing the core features users need for their job-to-be-done, or feeling like a countdown to a paywall rather than a useful product — produces neither. You get a small share converting to paid and no organic pull. That is the trap: a free tier built to demo the product instead of to be a product.
How to actually answer the question
The recommendation question is a heuristic, not a metric you can read off a chart. You can triangulate toward an answer by watching two things in parallel:
- LTV of installs
- Overall retention across both free and paid users
Opal used this pair when testing how much of its core 'blocks' feature to give away. Blocks are the scheduled restrictions that block distracting apps during specific times. The team tried everything from a single block, which is very restrictive, up to several, measuring both metrics at every step.
The target was the point where both metrics moved up together. That combination reflects the health of the whole user base, rather than paid conversion alone or the efficiency of the free-to-paid funnel.
Three blocks turned out to be the line. In Kenneth's account, three is enough for free users to get a great experience: they can really use the app and try a few different things, while power users who are convinced still pay because they want more. Three blocks gave non-payers enough utility to build a habit and recommend the app, and left enough headroom that committed users still had a reason to upgrade.
The diagnostic rule that follows: if LTV and retention do not both improve as you adjust the free tier, you are optimizing one segment at the other's expense.
What 'yes' actually unlocks
When Opal moved from a hard paywall to a genuinely generous free tier, pay penetration — the share of monthly active users on a paid plan — fell from 20% to 9%. That looks like a disaster on a conversion dashboard. It was not.
The free tier reached a segment Opal could not reach behind a paywall: high school and college students, who came to make up two-thirds of Opal's DAUs. Those users pulled pay penetration down, but they did something more valuable. They told their classmates, then their schools. Students recommending the app to administrators led to Opal for Schools, now a contracted B2B revenue line — a distribution channel that did not exist under the hard paywall.
Overall, the shift pushed Opal past one million daily active users. In Kenneth's words, it is a short-term, scary drop, but in the long term it pays back tenfold.
These are Opal's reported results, captured from the podcast conversation. They are a reference point, not a forecast for your product; the transferable part is the method, not the numbers.
The takeaway
If your non-payers would not recommend the product, you do not have freemium — you have a trial in disguise.
Adding more to the free tier is only one option. Another is to redistribute what is free: loosen one feature, tighten another, and reshape the package so non-payers get real value instead of a preview. Use the recommendation question to set direction, then track LTV and overall retention side by side to locate the boundary that fits your app.
Three questions to ask about your free tier:
- Does it unlock enough utility to build a habit or make progress toward the user's goal?
- Does it unlock enough to make them recommend the app?
- Does it keep enough back to give users a reason to upgrade?
Most teams tune the free tier to maximize conversion. The teams that win tune it to maximize recommendation and let conversion follow.
A quick decision checklist
Before you change a single limit, write down what you are testing and what would count as a pass.
- Name the free-tier lever you are changing (a count, a feature, a time window) and the values you will try.
- Define the job-to-be-done a free user must be able to finish. If the free tier cannot complete it, the test is about the wrong thing.
- Track LTV of installs and overall retention together, across free and paid users, at each setting.
- Treat a rise in one metric with a fall in the other as a fail, not a win.
- Ask the recommendation question directly of non-payers, and treat a 'no' as a signal to give away more or differently.
- Expect pay penetration to move. A drop is not automatically bad if reach and recommendation are rising.
A worked example using Opal's captured evidence: the lever was the number of blocks; the values ran from one to several; the pass condition was both LTV and retention rising; the result was three blocks. The same shape of test applies to any countable free-tier limit in your app.
Frequently asked questions
What is the recommendation test for a free tier?
Ask whether a non-paying user would recommend the app to a friend. If the answer is no, the free tier is not delivering enough value to fuel organic growth, and you need to give away more — or give away differently.
Why isn't conversion rate enough to judge a free tier?
Conversion rate, paid penetration and LTV all describe paying users. They say nothing about whether non-payers are an asset or dead weight, which is exactly what determines whether freemium produces word-of-mouth.
Which two metrics should I track in parallel?
LTV of installs and overall retention across both free and paid users. The free-tier setting you want is the one where both move up together; if one improves while the other falls, you are trading one segment against the other.
Does a generous free tier always mean lower revenue?
Not necessarily. Opal's pay penetration fell from 20% to 9% after moving from a hard paywall to a generous free tier, but the free tier reached students who became two-thirds of DAUs, recommended the app to schools, created a contracted B2B line, and helped push the app past one million daily active users. These are Opal's reported results, not a guarantee for other products.
Is the fix always to give away more?
No. Sometimes the fix is to give away differently — more of one feature, less of another — so the free experience is genuinely useful rather than a teaser. The recommendation question tells you whether you have a problem; the dual-metric test tells you where the line is.