Don't Spend Money on an App Yet: What a Business Really Needs

The hard truth first

I talk to founders every week. Many arrive with the same plan: "I have some savings, I have an idea for a ride app (or a food app, or a cleaning app), and I want to build it." They want a price and a timeline. Most of them do not need an app yet.

Here is the honest version. In the AI era, anyone can build software. Templates, white-label platforms, freelancers and AI coding tools have made getting a working app cheaper and faster than ever. Building the app is the easy part.

The hard part is the thing no developer can do for you: getting money out of customers' pockets and into yours, again and again, at a profit. If you cannot do that by hand, an app will not do it for you. It will only help you lose money faster and more professionally.

So before you spend a single dollar on development, answer these questions honestly:

• Who exactly will pay you, and have any of them already paid?
• Where will your drivers, restaurants or service providers come from, and why will they stay?
• Who will answer the phone at 11 p.m. when an order goes wrong?
• Do you have a plan, and the money, to survive the first six months after launch?

If you cannot answer these, this article is for you.

Why "just build the app" fails

CB Insights studies why venture-backed startups shut down. Its March 2026 analysis looked at 431 companies that closed since 2023. Where a reason could be identified, about 70% ran out of capital and 43% had poor product-market fit. Bad timing and unsustainable unit economics followed. Almost none of these failures were about the code.

Read those two numbers together. Companies run out of money because they spend before they know people want what they sell, or before they know each sale makes money. An app is often where that money goes first.

Three patterns I see again and again:

1. The app is finished, but nobody is on the other side. A delivery app with no riders, or a ride app with no drivers, is an empty shop with a beautiful sign.
2. The launch budget was the build budget. The founder spent everything on development and has nothing left for marketing, incentives or support. Launch day comes and goes with silence.
3. The founder has never sold the service manually. They do not know the real price customers accept, the real cost per order, or the real problems in delivery. They are guessing, and the app hard-codes the guesses.

Validate demand before you build

Validation means collecting evidence that people will pay, not opinions that they might.

Talk to customers about what they do now

Ask about past behaviour, not your idea. "How did you order lunch last week? What went wrong?" tells you far more than "Would you use my app?" People are polite. Compliments cost them nothing.

Pre-sell

The strongest proof is money. Ask for a deposit, a prepaid package, a monthly subscription paid up front, or a signed letter of intent from a business customer. If nobody will commit even a small amount, that is your answer, and it cost you almost nothing to get it.

Run a manual pilot (a "concierge MVP")

Do the service by hand for a few weeks in one small area. Take orders on WhatsApp, a phone line, a Facebook page or a simple form. Match them to drivers or providers yourself. Track every order in a spreadsheet.

This is not a new idea. Some of the largest on-demand companies started exactly this way:

• DoorDash began when its founders built a simple page called PaloAltoDelivery.com in one afternoon. The Stanford Daily reported that they got their first order right away and delivered a couple of hundred orders themselves before joining Y Combinator.
• Gojek started in Indonesia around 2010 as a small call centre with about 20 motorbike taxi drivers. Its app did not launch until January 2015.

Paul Graham's well-known essay "Do Things That Don't Scale" makes the same point for founders in general: early users rarely come to you on their own. "You have to go out and get them."

A manual pilot tells you your real conversion rate, average order value, delivery time, cancellation rate, cost per order and the questions customers actually ask. That is the data you need to scope an app properly.

Lock in the supply side

Every on-demand business is a marketplace. Customers are only half of it. The other half is drivers, riders, restaurants, cleaners or technicians, and they are often harder to win and keep.

Before you build, find out:

• How many providers you can sign up in your first area, and what it costs to recruit each one.
• What they earn now, and whether your commission still leaves them better off.
• What documents, licences, vehicles or training they need.
• Whether they will actually show up at peak times when you need them.

If you cannot sign up 20 reliable providers by hand, an app will not fix that.

Operations: the business behind the screen

An app shows a button that says "Order". Everything after that tap is operations:

• Dispatch and quality. Who checks that the driver arrived, the food was hot, the cleaner did the job?
• Customer support. Refunds, complaints, lost items, wrong addresses, disputes between customers and providers.
• Payments and payouts. Collecting money, paying providers on time, handling cash where cash is common, reconciling every day.
• Safety and compliance. Background checks, insurance, and the licences your city requires.

Write down who does each of these on day one. If the answer to every line is "me", you are not ready to scale, and you need to budget for help.

Cash flow: how money moves through your business

Marketplaces have tricky cash flow. You may collect money from customers on one schedule, pay providers on another, and pay payment fees, promotions and refunds in between. Many founders discover too late that growth makes the cash problem worse, because every new order needs incentives and support before it earns anything.

The US Small Business Administration's startup-cost guidance is a good, simple model even outside the US. It separates one-time expenses from monthly expenses, recommends planning monthly costs for at least one year, and shows a basic break-even formula: fixed costs divided by (price minus variable costs). It also suggests adding a buffer of about 10% for the unexpected. Do this exercise before you talk to any developer.

The full budget: the app is a fraction

When founders ask me "how much does the app cost?", I ask them to fill in the rest of this table first. The rows matter more than the numbers.

For context on the build line: at Joldisoft a white-label platform starts from US$3,000 and a custom MVP starts from US$10,000. If your total budget is close to that number, you do not have an app budget problem. You have a business budget problem, and spending it all on software will not solve it.

Red flags: stop and rethink

Pause before you build if any of these are true:

• "Everyone I asked said they would use it." Opinions are not orders. Ask for money.
• "We will figure out marketing after launch." Launch is when you need marketing most.
• "I'll copy Uber's features." Uber has thousands of engineers. You need the smallest product that runs your core loop.
• "The app will run itself." No marketplace runs itself in its first year.
• "My whole budget goes to development." Then you have no money to get customers.
• "I don't know the licences yet." Transport, food and payment rules can change your entire model. Find out first.
• "My co-founder will handle the business side" and nobody owns sales. Somebody must be accountable for revenue, by name.

Your pre-build checklist

Do not sign a development contract until you can tick most of these:

• I have spoken to at least 20 potential customers about how they solve this problem today.
• At least some customers have paid me, prepaid, or signed a letter of intent.
• I have run the service manually (WhatsApp, phone, spreadsheet) for at least a few weeks.
• I know my real average order value, cost per order and contribution margin from the pilot.
• I have signed up enough providers to serve my first area reliably.
• I know which licences, insurance and registrations my city requires.
• I have a written sales and marketing plan for the first 90 days, with a budget.
• I have budgeted for operations, support, servers, legal and maintenance, not just the build.
• I have enough runway to survive at least six months after launch with low revenue.
• One named person is accountable for running the business every day.
• I have scoped the smallest app that runs the core loop, nothing more.

When it is time to build

Once you have proof, the app becomes a tool that multiplies something that already works. That is the right time to spend. You will also spend less, because you know exactly what to build and what to leave out.

I will be direct about how we work. Joldisoft does not build for everyone. We say no to projects where the client's money is likely to be wasted. We prefer to work with founders who have validated demand, a sales and marketing plan, a budget for operations and runway, and someone accountable for running the business. We want long-term partners whose businesses survive, not one-off projects that end in an abandoned app. If that sounds like you, we are happy to talk. If it does not yet, use the checklist above first. It is free, and it is the most valuable work you will do this year.