When a club buys a player for 20 million, that isn't what the player costs. There's the salary over five years, the agent's commission, the signing bonus, the match bonuses. The figure in the papers is the visible part, usually the smallest one.
Software quotes work the same way. That's why you get three proposals ranging from 400,000 to 8 million FCFA for what you believe is the same product.
I'm going to give real numbers. Not cautious ranges: what I've seen practised on the Cameroonian market, and what the projects I've worked on actually cost.
The four tiers, with prices
300,000 to 800,000 FCFA — the storefront. A site or app that presents your business. Fixed pages, a contact form, maybe a catalogue you can't edit. No user accounts, no data that changes. It's honest work, and it's enough for many businesses.
The trap: it gets sold to you as "an app". If you can't change the content yourself, it isn't software, it's a brochure.
1.5 to 4 million FCFA — the app that manages something. User accounts, data that gets created and edited, a back office to administer it. This is where most real needs sit: an ordering app, customer tracking, a living catalogue.
5 to 12 million FCFA — the app that takes money. The moment money is involved, the price changes in kind. Not because the "Pay" button is hard: because everything around it is. I've written a whole article on this — the six states to model, the webhooks, the reconciliation. Budget two to three weeks of work for a single operator's Mobile Money integration, done properly.
15 million FCFA and up — the platform. Several user types with different permissions, several applications talking to each other, multi-tenancy, serious business rules. Glyph — a print marketplace with a Go backend, a mobile app and an admin console — sits here. So does Driivio: a consumer app, a role-based back office, a white-label foundation.
What actually moves the price
It's almost never the number of screens. It's five things.
1. The number of roles
An app with one type of user is a product. With four — customer, employee, manager, administrator — it's four products sharing a database. Each role has its screens, its permissions, its edge cases.
Driivio has the school, the instructor, the partner and the administration. It isn't four times the price of a single role, but it's well more than double.
2. Money
A product that takes payments costs structurally more, and the difference isn't in the development: it's in everything that has to be planned for when it fails. An audit log, a refund procedure, a "needs checking" state and someone to handle it.
A supplier quoting three days for a payment integration hasn't planned for that. You'll pay the difference later, in weekends spent refunding people by hand.
3. Offline mode
An app that must work without a network costs 30 to 40% more on the features concerned. A write queue, conflict resolution, local database migrations.
This isn't decorative on the Cameroonian market. I've explained why here. But it's a cost, and it should be decided deliberately rather than discovered.
4. Who owns the data
An app wired to a third-party service costs less up front. An app with its own backend costs more, and belongs to you.
The question isn't "what does it cost" but "what happens if that supplier disappears". I watched my own database become unreachable for nine days. It changes how you look at a dependency.
5. What comes after delivery
This is the line item every quote forgets, and the one that costs most over three years.
A living app needs hosting (15,000 to 80,000 FCFA a month depending on load), the updates Apple and Google impose — at least one a year, or you're pulled from the stores — fixes, and evolution.
Budget 15 to 25% of the initial cost per year. A quote that doesn't mention this isn't a quote, it's a deposit.
The five questions to ask before signing
These cost nothing and spare you most of the bad surprises.
"Who owns the source code?" If the answer isn't "you", you're renting your product without knowing it. Insist on a Git repository in your name from day one, not on delivery.
"What happens if you disappear tomorrow?" A good answer contains the words documentation and README. A bad one is "that won't happen". Half my engagements are picking up a project whose developer left without writing anything down.
"What is NOT in this price?" A serious supplier has an answer ready. This is the question that reveals the most.
"What does year two cost?" See above.
"Show me something you shipped that's still running." Not a mockup. A URL, an app in a store, a client I can call.
The market's underlying problem
There's an asymmetry here that hurts everyone.
The client can't evaluate a technical proposal — that's normal, it isn't their job. So they compare the one thing they understand: the price. The cheapest quote wins.
The supplier knows this. So they quote low, cut what isn't visible — tests, error handling, documentation — and deliver something that works in the demo. Six months later the product is unmaintainable, the client is disappointed, and they'll tell everyone that "developers are a scam".
That cycle damages the whole profession, and it feeds on silence about prices. Which is exactly why this article has numbers in it.
A low quote isn't a good deal. It's a debt whose due date you don't know yet.
What I actually do
I scope before I price. One or two hours of free conversation, during which I'm mainly hunting for what will be expensive and that you haven't seen: the hidden roles, the edge case that structures everything, the point where money moves.
Then I propose a scope, not a single number: what's essential to go live, and what can wait for the next version. Almost always, half the initial list can wait — and that half is your year-two budget.
If you have a project and you want an honest estimate rather than a number designed to win you over, message me on WhatsApp. The first conversation is free, and sometimes I come out of it telling you that you don't need an app at all.