Which Monetization Model Actually Fits Your Project
In short. There are not many models and all of them work — for somebody. The question is never which is best, it is which one matches the traffic you have and the content you have. Subscriptions on incidental search traffic do not pay back; one-off sales to a loyal audience leave money on the table; advertising at small volume does not pay at all. Below: how to choose using two facts you already know about your own project, and why combining models usually subtracts rather than adds.
This page is about choosing. If the model is already set and the question is how to get more out of it, that is a different page.
The two facts that decide it
Everything collapses into two questions about your project, and you already know the answers.
-
Do people come back on their own? Not "could they" — do they, without reminders and without paid retargeting. If yes, you have a relationship and it can be sold to repeatedly. If no, every sale is a first sale, however much you would prefer otherwise.
-
Does the content keep arriving? Regularly and predictably — or is it finite and mostly already made?
| Content keeps arriving | Content is finite |
|---|
| People return |
Subscription. The only cell where it honestly works |
One-off purchase plus upsells. A subscription collapses when the archive runs out |
| People do not return |
Advertising and affiliate revenue: monetise the flow, not the person |
One-off sale or affiliate revenue. The hardest cell of the four |
⚠ The characteristic mistake of this industry is selling a subscription from the bottom row. Subscriptions look attractive because they produce predictable revenue, but the predictability comes from people returning, not from the billing method. If they do not return by themselves, a subscription is simply a one-off sale with high churn and refunds attached.
What each model demands of you
| Model | What it must deliver | What you pay with | Where it breaks |
|---|
| Subscription |
New material on a schedule |
A publishing commitment and constant churn work |
Publishing stops; cancellations arrive quietly, unannounced |
| One-off purchase |
Obvious value in a specific item |
Every sale made from scratch |
No repeat revenue, so acquisition eats the margin |
| Affiliate revenue |
Traffic volume and honest tracking |
Dependence on somebody else's terms and payouts |
The partner changes rates or drops the category |
| Advertising |
Impressions, in quantity |
Attention that goes somewhere other than your offer |
Pays nothing at small volume; competes with your own sale |
Selling access to an audience
(listings, profile placement) |
Demand from visitors |
You sell placement, not outcomes |
Too few visitors and advertisers stop renewing, visibly and fast |
Why "all of them at once" rarely works
The temptation is understandable: subscription plus one-off sales plus advertising plus affiliate links. On paper it adds. In practice it subtracts, for three reasons.
-
The models compete for the same attention. A partner's banner on the page where you sell a subscription removes the visitor at precisely the moment they were closest to paying.
-
They compete for the same content. Whatever goes into the free area to earn impressions can no longer be sold to a member.
-
Each demands its own work. Subscriptions demand publishing, advertising demands volume, affiliates demand tracking. Three models at half effort lose to one done properly.
The combination that does work is usually a single pairing: a primary model, plus one secondary that monetises the people who were never going to use the primary. Affiliate revenue from visitors who are leaving, advertising in the free area — but never over the payment page.
Checking the choice against numbers
Value per visitor = revenue over a period ÷ visitors over the same period
The model holds while: value per visitor > cost to acquire one
Count it per traffic source rather than in aggregate: the average is exactly what hides the fact that one channel feeds the project while another consumes it. And count over a horizon that matches the model — a three-month cohort for subscriptions, the sale plus repeats over the same period for one-off purchases.
If value per visitor is below acquisition cost under every model, the model is not the problem. Either the traffic is wrong or the thing being sold is.
Signs the wrong model was chosen
- There is a subscription, nothing new has been published for a month, and nobody noticed.
- Revenue rises only with ad spend and falls with it.
- More than half the income comes from a single partner.
- The dispute ratio is above the neighbourhood average, usually a sign that people did not understand what they were paying for.
- The free area is so good there is no reason to pay — or so thin there is no reason to believe paying is worth it.
- There are other people's links on the payment page.
What to check right now
- Which cell of the table you are actually in, rather than the one you designed for.
- Whether people return unprompted — measured, not assumed.
- Whether you can hold a publishing schedule for twelve months, if you chose a subscription.
- Whether value per visitor is calculated per channel.
- Whether a secondary model is competing with the primary on the same page.
- What share of revenue depends on one external partner.
How adults.dev approaches it
We start from the two questions at the top and from measurement, not from a feature discussion. Fairly often a project turns out to sit in one cell while being built for another, and that explains the numbers better than any amount of additional development would.
What gets built after that is whatever the chosen model actually requires: cohort reporting, retry logic and visible churn for a subscription; a short path to payment and upsells for one-off sales; tracking that survives changing partner for affiliate revenue. One model built completely beats four built halfway.
Talk through your model:
Telegram: @adultsdev
E-mail: info@adults.dev