Anyone Can Build an App Now. Can You Sell It?

The uncomfortable truth

I run a company that builds apps. So it may sound strange that the most useful thing I can tell you is this: the app is the easy part.

Ten years ago, building a ride-hailing or food delivery platform was a serious engineering project. Today, AI coding assistants write a large share of routine code, ready-made white-label platforms exist for almost every marketplace model, and no-code tools let a non-technical person ship a working prototype in a weekend. Software has not become free, but it has become cheap enough that "we have an app" is no longer an advantage. Your competitor can have the same app next month.

What has not become cheaper is the hard part: getting customers to pay you, getting restaurants, drivers or service providers to work with you, and running the operation every single day without it falling apart. That is what this article is about. No motivation, no hype. Just the questions that decide whether your money turns into a business or disappears.

What AI has actually changed

Let us be precise, because there is a lot of noise.

• Most developers now use AI tools. In the Stack Overflow Developer Survey 2025, 84% of respondents said they use or plan to use AI tools in their development process, up from 76% the year before, and about half of professional developers use them daily.
• On well-defined tasks, the speed-up is real. In a controlled experiment published by researchers from Microsoft, GitHub and MIT, developers with GitHub Copilot completed a standard task (writing an HTTP server in JavaScript) about 56% faster than those without it.
• It is not magic on complex, real-world code. A 2025 study by METR found that experienced open-source developers working on their own large projects actually took 19% longer with AI tools, while believing they had been faster. The same survey above found more developers distrust the accuracy of AI output than trust it.

So the honest picture is: building has become faster and cheaper, especially for standard features, but quality still depends on people who know what they are doing. Either way, the cost of producing software is falling. When something becomes cheap and widely available, it stops being the thing that separates winners from losers.

Why startups really die

CB Insights has studied startup post-mortems for years. Its 2026 analysis of 431 venture-backed companies that shut down since 2023 found that 70% ran out of capital and 43% suffered from poor product-market fit; 19% had unsustainable unit economics. CB Insights itself points out that running out of money is usually the final cause, not the root one. Companies run out of money because not enough customers paid them enough, for long enough.

Notice what is not on top of that list: "the app had bugs" or "the technology was not good enough." Founders rarely fail because they could not build. They fail because they could not sell, or could not sell profitably.

The real moats in an on-demand business

If the app is not your moat, what is? In ride-hailing, food delivery and home services, it is almost always some combination of these:

Read the right-hand column again. Almost everything that decides your survival is human work.

The founder's real job

Here is the plain version of a founder's job, without the inspirational language: move money from customers' pockets into the company, and do it in a way that costs less than it brings in.

Everything else serves that. Product decisions, hiring, design, technology: all of it exists to make that transfer happen more often, at a better margin. If you are not comfortable asking people for money, negotiating with suppliers, and hearing "no" twenty times before one "yes", you do not yet have a business. You have a project.

That is not an insult. Many good businesses started with founders who learned to sell. But they learned it before they spent their savings on software, not after.

How the successful ones started

Paul Graham's well-known 2013 essay Do Things That Don't Scale makes the case that startups take off because founders make them take off, often through manual, unglamorous work. He describes the Airbnb founders going door to door in New York to recruit hosts, and the Stripe founders setting users up on their own laptops on the spot instead of sending a link.

DoorDash is an even closer example for anyone in delivery. According to the Stanford Daily in 2013, its founders built a basic site called PaloAltoDelivery.com in an afternoon, got an order almost immediately, and delivered a couple of hundred orders themselves to Stanford before joining Y Combinator. The "technology" was simple. The work was the deliveries.

The lesson is not "build less." It is "sell first, then build what the selling has proved."

How to test your sales ability before you build

You can test most of your business with a phone, a spreadsheet, a WhatsApp group and a few weeks of effort. Here is what real evidence looks like.

1. Pre-orders or deposits

Ask early customers to pay something before the service is fully live: a prepaid bundle of deliveries, a discounted first month of a cleaning subscription, a corporate account deposit. People say yes to surveys. They only say yes with money when they mean it.

2. Letters of intent from the supply side

For a marketplace, supply is half the business. Ask restaurants, fleet owners or service providers to sign a short, non-binding letter of intent stating the commission or fee they are willing to accept and when they would start. A signature is a small step, but it separates "sounds interesting" from "I will do this".

3. A manual pilot

Run the service by hand for a few weeks in one small area. Take orders by phone, WhatsApp or a simple form. Dispatch drivers from a group chat. Track everything in a spreadsheet: orders, delivery times, complaints, refunds, repeat customers, cost per order.

This tells you things no app can: whether customers reorder, which hours are busy, how often things go wrong, and whether the numbers work when you count every cost.

4. Signed supply, not promised supply

Before launch, you want named restaurants, drivers or technicians who have agreed to start on a specific date. "My cousin knows twenty drivers" is not supply. Twenty drivers with their documents collected and a start date is supply.

5. One repeatable acquisition channel

Find at least one way to get customers that you can repeat and measure: an office building, a university, a neighbourhood community group, a partner business. Know roughly what each customer costs you to acquire. If you cannot name a channel, you do not yet have a marketing plan.

The checklist: are you ready to build?

Tick these honestly before you sign any development contract, with us or anyone else.

• I have spoken to at least a few dozen potential customers about their problem, not about my idea.
• Some of them have paid, pre-paid or put down a deposit.
• I have signed letters of intent from restaurants, drivers or service providers.
• I have run the service manually, even for a few days, and I have the numbers.
• I know my average order value, my commission and my cost per order.
• I have one acquisition channel I have tested and can repeat.
• I have a budget for marketing and operations after launch, not only for the app.
• I know who will handle support, onboarding and disputes every day.
• I can survive several months of slow growth without shutting down.

If you ticked most of these, you are in a strong position, and software will multiply what you already have. If you ticked few, the best investment right now is not an app. It is a few weeks of selling.

Hard questions to answer before you spend money

These are the questions I ask founders on our first call. They are not meant to discourage you. They are meant to stop you from paying for an app that nobody uses.

1. Who exactly pays you, and why would they switch from what they use today? "Everyone who eats food" is not an answer.
2. What do you do better than the big players in your city? Cheaper, faster, a niche they ignore, a community they do not serve?
3. How will the first 100 customers find you? Name the channel and the cost.
4. How will the first 20 suppliers join, and why would they stay?
5. What happens in month three, when the launch excitement is gone?
6. Who answers the phone at 10 pm when an order goes wrong?
7. How long can you keep going if growth is slower than you hoped?

If you cannot answer these yet, that is fine. It just means your next step is research and selling, not development.

Where an app genuinely helps

None of this means apps do not matter. Once you have customers and supply, software is what lets you grow without drowning: automatic dispatch, live tracking, digital payments, ratings, payouts, reporting. A good platform turns a manual operation that works for 30 orders a day into one that works for 3,000.

That is also why we offer a fast, affordable starting point. A white-label platform from Joldisoft starts from $3,000 and can be live in two to three weeks, so you spend your money on selling and operating rather than on reinventing an ordering screen. When the business is proven and you need something unique, an MVP or custom build makes sense. We built our model around one belief: software should multiply a business that already works, not replace the work of building one.

The bottom line

In the AI era, anyone can build software. That is good news, because it makes testing ideas cheaper. It is also a warning, because your app will not protect you. The founders who survive are the ones who can sell, recruit supply, control costs and show up every day to run the operation.

So before you ask "how much does the app cost?", ask a harder question: can I sell this? Then go and prove it.