How do free apps make money? Here's how ours does.

Every free app is paid for somehow. If you're a user, it's worth knowing by whom. If you're building one, it's the decision that shapes everything else.

Here are the five models, what they need to work, and the one we picked for MeltsWell, with the actual prices.

The five ways

1. Ads

You're free, advertisers pay to reach your users. Works at scale: millions of daily users, or a very specific audience someone pays a lot to reach. Below that, ad revenue on a small app is pocket change and the ads make the product worse.

Needs: huge volume, or a niche with expensive buyers.

2. Freemium and subscriptions

Free tier does the basic job. Paid tier does more. Most software you pay for works like this. It works when the free tier is genuinely useful (so people stay) and the paid tier solves a problem the free user hits on their own (so they upgrade without being nagged).

Needs: a clear line between "enough" and "more," and users who make money or save time with the paid version.

3. Commission, or take rate

Free to use, the app takes a cut of every transaction. Marketplaces, delivery, payments. Airbnb, Uber, Stripe. Works when the app is in the middle of the money and adds enough that people accept the cut.

Needs: payments inside the app, and volume, because you earn a percentage of small numbers.

4. Paid placement

Free to list, pay to be seen more. Featured spots, promoted listings, sponsored results. Yelp, app stores, most directories.

Needs: enough eyeballs that being featured is worth paying for.

5. Selling data

Free because the user is the product. We're not covering this one because we don't do it and you shouldn't either.

What MeltsWell does

MeltsWell is a marketplace for home bakers and small sweet shops. Buyers find creators on a leaderboard, look at their page, and order on WhatsApp or Instagram.

We run 2 and 4, and deliberately not 3.

Free, forever. Any creator gets a page with their top 3 creations, a shareable link, one coupon, and a place on the leaderboard. No card, no expiry. Every new creator also gets 14 days of everything, so they see what the paid plans do before the free plan kicks in.

Starter, $7.99 a month. Verified badge, 15 products plus 5 special orders, analytics for the last 15 days. Launch price $5.99, or $3.99 a month paid yearly.

Growth, $13.99 a month. 30 products, full analytics since day one, and the thing that sells it: you see who's reaching out, which products get clicked, what converts. Launch price $10.49, or $6.99 yearly.

Featured, $9.99 one time. Thirty days on the homepage. No subscription. Any live page can buy it from the dashboard.

Commission: zero. Orders happen on WhatsApp between the buyer and the baker. We never touch the money.

Why no commission

A take rate is the obvious model for a marketplace (how to build a marketplace app covers the rest of the build). We said no for three reasons.

We're not in the payment. Moroccan home bakers already take orders on WhatsApp and get paid in cash or by transfer. Forcing that through our checkout to skim 10% would make the product worse for both sides and most of them would just… keep using WhatsApp.

The numbers are small. A cake is $20. Ten percent is $2. To make $1,000 a month we'd need 500 orders to flow through us. A $7.99 subscription needs 125 subscribers, and those people get value every day, not per order.

Trust. "No marketplace fees, no middleman" is the first thing a baker asks about and the first thing we can say yes to. That line has closed more signups than any feature.

Why a free tier at all

The first pricing spec, written in May, had no free tier. Two paid plans, $14.99 and $29.99, and a three-month trial. We rewrote it in July before launch: free forever, and paid prices cut in half.

The reason is the leaderboard. A marketplace with 6 bakers is not a marketplace. The free tier is what gets a page for every baker in a city, and the leaderboard only works when that's true. The paid tiers pay for it. The free tier is the product.

What upgrades people

Not the product limit. Analytics. The free plan shows nothing. Starter shows 15 days of views. Growth shows who tapped WhatsApp, which product they clicked, and the conversion rate. A baker who sees 200 views and doesn't know if anyone reached out will pay $10 to find out. That's the whole upgrade path, and it's honest: we're selling information they want, not removing something they had.

Where we are

59 creators across 18 cities as of mid-September. Launch pricing runs through the promo window. I'm not publishing revenue yet; the roster is too small for the number to mean anything. When it's a real number, it'll be on this blog.

If you're picking a model

• Under 10,000 users: forget ads.
• You're not in the payment flow: forget commission. Don't force yourself into it just to charge.
• Your users make money with your app: subscriptions. Price against what they earn, not against what it costs you.
• You have a feed, list, or ranking: paid placement is nearly free to add and people understand it instantly.
• Free tier must be useful alone. A free tier that's a demo isn't a free tier, it's a trial with no end date, and people know.

We're building for founders too, and pricing is the first thing we talk about on the scoping call, before a line of code. If you're at that stage, tell us what you're building.