The most common question I see from business owners is also the one that gets the worst answers. Someone posts "I have an app idea, where do I start?" and within an hour there are six replies that say "DM me." None of them say where to start. They say where to start paying.
Here is the order I would actually go in, and it costs close to nothing until step four.
Key Takeaways
- Write down what people do today without the app. That tells you what the app replaces.
- Name every kind of person who will log in. Each one adds a real chunk of work.
- Pick the one thing a user must be able to do, and build only toward that first.
- Only then choose how to build it: an AI builder, one developer, or a small team.
Step One: Describe What Happens Today
Before the idea, the current mess. Is the job done today in a spreadsheet, a WhatsApp group, a paper notebook, a phone call, or not at all?
This matters more than anything you imagine the app doing. If the work happens in a spreadsheet today, the app is a better spreadsheet, and that is a well-understood kind of software. If nothing happens today because nobody does the job yet, you are building a new behavior, and the hard part is getting people to change what they do. No developer can fix that for you.
Write three or four sentences. "Customers call me to book. I write it in a paper diary. Twice a month I double-book someone and lose the job." That paragraph is worth more than a logo.
Step Two: List Who Logs In
Customers. Your staff. You. A second business on the other side, like a supplier or a service provider.
Count them. One kind of user is a simple app. Two kinds means two sets of screens, two sets of permissions, and a whole class of bugs where one person sees something meant for the other. A marketplace, where buyers and sellers both log in and money moves between them, is the most expensive shape a small idea can take.
Most owners discover at this step that their idea is bigger than it sounded. Better to find out now, for free.
Step Three: Find the One Thing
Finish this sentence: "This is worth it if a user can ______."
Book a slot. Pay a deposit. See where the technician is. Get a quote in under a minute. Pick one. Everything else goes on a list called "later," and you will be surprised how much of it stays there. The first working version should do that one thing well and almost nothing else.
Step Four: Choose How To Build It
Now, and only now, does money come in. There are three honest paths, and the price ranges below are the same ones our free honest read tool uses.
| Path | Typical cost | Typical time | Where it breaks |
|---|---|---|---|
| Do it yourself with an AI builder | $0 to $200 a month in tool fees | days to a few weeks | two kinds of logins, payments, data you cannot lose |
| One developer for a few weeks | $3,000 to $12,000 fixed | three to eight weeks | no second pair of eyes, nobody maintains it later |
| A small team at a fixed price | $15,000 to $45,000, paid by milestone | six to twelve weeks | the plan growing after the price was agreed |
The answers from steps one to three pick the row for you. One kind of user, no payments, nothing you would cry about losing: start with a tool. Two kinds of users, or data your business depends on, usually means a developer. Add money moving between people and you are usually in the bottom row, however simple the screens look.
A Concrete Version
Say you run a mobile dog-grooming business with two vans. Today, bookings come in by text and live in a shared Google Calendar. Twice a month a van shows up at the wrong address.
Step one: the app replaces the texts and the calendar. Step two: customers book, and two groomers check their day. That is two kinds of users. Step three: "worth it if a customer can book a slot that is actually free, at the right address." Step four: the calendar has to stay true for the customers and both vans at once, and you want a deposit to cut no-shows. Money moves. That lands in the small-team row, likely toward the lower end of $15,000 to $45,000, because the core is one booking flow done carefully.
If you drop the deposit and only need customers to request a slot you confirm by hand, the same idea becomes a form and a calendar. An AI builder can do that this weekend.
The Honest Counterpoint
This order assumes the idea is to make an existing business run better. If you are starting a business whose product is the app, steps one to three are harder, because there is no "today" to describe. In that case the first thing to build is often no software at all: a landing page, a waitlist, or doing the service by hand for ten customers. Paying for code before anyone has asked for the thing is the most expensive way to learn it was the wrong thing.
Frequently Asked Questions
Do I need to know how to code to start?
No. The first three steps are about your business, and you know it better than any developer. You need a developer at step four, and even then a good one will ask you the questions above before quoting. See questions a good software developer should ask you.
Should I get an NDA before I describe my idea?
Most developers will not sign one for a first conversation, and the idea is rarely the valuable part. How you run it is. Our post on how to protect your idea when hiring a developer covers what actually protects you.
How do I know if the quote I get is fair?
Compare it to the three rows above for the shape of your idea, and read what a good software quote looks like. A quote with no written list of what is included is not a quote yet.
What if I cannot afford the row my idea lands in?
Cut the idea until it fits a cheaper row. Remove the second kind of user, or handle payments outside the app for now. Our guide on what to leave out of the first version of your app walks through how.
The Bottom Line
Describe today, count who logs in, pick the one thing, and only then pick how to build. If you would rather have someone walk those questions with you, the honest read does it in a few minutes, free, and sometimes the answer is that you do not need a developer yet. When you do, our software factory builds at a fixed price.
Roberto Espinoza is CEO of Ruzora, which builds custom software for business owners at a fixed price and places pre-vetted senior LATAM engineers with US teams. Get a free honest read on your idea.
