There is a particular kind of business that only exists for a while. Someone notices that a hard thing has become easy, builds a service around it, and for a few years sells it to people who have not noticed yet. Then everyone notices, and the window closes.
Building mobile apps is in that window right now. It used to need two developers and three months. It now needs a platform and a conversation. The people who will own that market in 2028 are signing their first clients this year.
What actually changed
Not the apps. A native app is the same amount of engineering it always was. What changed is who has to do it.
An AI assistant can now take a plain-language request — “put the shop in a bottom tab, dark theme, hide the cookie banner” — and turn it into a configured, signed, store-ready app. Not a mockup. A build you can install. The client describes, the platform assembles, you take the money.
That is the whole business. You are not a developer in it. You are the person who owns the machine and knows who needs one.
Your business gets smarter while you sleep
This is the part people miss, and it is the strongest argument for starting now rather than later.
In most businesses, improvement costs you something. Better service means hiring. Better product means building. Here, the main component — the AI that configures the apps — is improving on someone else’s budget, at a pace nobody in this industry has matched before.
Every time the models get better at understanding a vague request, every client on your platform has a slightly easier time, and your support load drops slightly. You did nothing. You will do nothing next time either.
The same is true of the platform itself. Native storefronts for WooCommerce and Shopify did not exist in this product three months ago; they do now, and every existing licence holder got them. Your clients’ apps gained a feature you never commissioned.
Compare that with a business whose core you own outright. A web studio with five developers improves at the speed of five developers. This improves at the speed of the whole field.
What it costs to be in it
Less than people assume, and the shape of the cost is unusual.
- The platform: bought once. A self-hosted licence on CodeCanyon, paid one time. Not per app, not per client, not per month.
- Hosting: three months free on the ready-made install, then the price of ordinary shared hosting. There is nothing to configure — you enter your domain, the server deploys itself, and an email tells you where to point the DNS.
- A domain. Yours, bought at any registrar, for about the price of lunch.
That is the list. No per-app fee, no revenue share, no platform tax that grows as you grow. The thirtieth client costs you almost exactly what the third one did.
It is worth seeing what the alternative looks like. AppInstitute publishes $599 a month for its reseller tier. GoodBarber publishes €280 a month, or a €12,000 one-time buy-out of the same hosted platform — which is the clearest statement anyone in this market makes about what the monthly fee really compounds to.
Those are fair prices for what they include: their servers, their support desk, their uptime. But they are a tax on your growth, and they are charged hardest in month two, when you have one demo app and nobody has called back.
What is still being built
Being straight about this: ready-made landing pages for your own storefront are in development, not shipped. The idea is that you pick a design for the site where your clients sign up, instead of commissioning one. When it lands, existing licence holders get it like everything else.
Today you still put your own front page together. It is the one piece of work the platform does not do for you yet.
The clock: the part nobody tells you
Here is the honest reason to start early, and it has nothing to do with motivation.
Search engines do not trust new sites. Not because of anything you did — they simply have no reason to spend attention on a domain nobody links to. Everything about getting found takes a long time, and none of it can be bought back later.
We can be specific, because this site is the experiment. Over three weeks we published thirteen articles, all of them written rather than spun, all technically clean — correct canonicals, a valid sitemap, structured data, fast responses.
| What happened | How long it took |
|---|---|
| Google crawled the site at all | about 9–13 pages a day, total |
| Articles indexed without being submitted by hand | roughly two of thirteen |
| First impression for a non-brand search | around two weeks after indexing |
| Where those searches ranked | positions 74–98 |
| Clicks from them | none yet |
Read that table again, because it is the most useful thing in this article. A clean site with real content, published steadily, took three weeks to reach the third page of results for its own subject. That is not failure — that is the normal shape of the curve, and anyone telling you otherwise is selling something.
The curve goes up. But it starts when you start. A competitor who began a year before you is not a year ahead in effort; they are a year ahead in trust, and trust is the one input you cannot buy, borrow or accelerate.
Which is why the cheapest thing you can do today is the thing that only time pays off: get the domain, put something real on it, and let the clock run while you learn the rest.
Claim a niche while they are open
The market is not one market. “App builder” is brutally competitive; “apps for driving schools in Portugal” is empty.
The people doing well at this do not compete on being the best platform. They compete on knowing one trade: they speak restaurant, or gym, or clinic, and they have five apps in that trade to show. The second client in a niche is dramatically cheaper to win than the first, because the pitch, the demo and the objections are already done.
There are a lot of trades and a lot of countries, and almost none of them are taken. Pick one you already understand — what you used to do, where you used to work, the language you speak that your competitors do not.
What this is not
The usual section, because an article that only sells is worth nothing.
- It is not passive. The platform builds the apps. It does not find the clients. Everything hard about this business is on the sales side, and it stays there.
- It is not instant. See the table above. If you need revenue this quarter, this is the wrong shape of business.
- You run a server. Self-hosted means updates, backups and the occasional 3am problem are yours. That is the price of the economics; if you would rather pay someone to carry it, rent a platform instead and accept the monthly fee.
- Not every client is a good client. Businesses whose customers never return should not buy an app, and selling them one produces a refund and a bad review.
Where to start, concretely
- Pick the trade you already understand. One, not three.
- Buy a domain. Today, not after the plan is finished.
- Get the licence and the install — the hosting is free for the first three months, so the domain is genuinely the only money at risk while you look around.
- Build two apps for businesses you know, at cost, with permission to show them.
- Put a real page on the domain and start writing about the trade you chose. The clock starts here, and it never starts earlier than this.
The honest summary: the tools got good enough that one person can run this, the economics work because the platform is bought rather than rented, and the only genuinely scarce input is time already spent. Everything else is available to anyone who decides to start.
The economics in detail — rent versus own, where the break-even really sits — are in rent the platform or own it. Who to sell to is in which small businesses actually buy an app, and what to charge them is in what to charge a small business for their app.