How do I turn a problem I see every day into a tech product?
By Bonaventure Ogeto, Founder and CEO, Mctaba Labs
Published
Write the problem down for one specific person, map how they handle it today step by step, and find the single step that costs them most. Deliver that step by hand for a few paying customers, then build software that does only that. A product is a problem narrowed until one small tool can solve it.
You see it every morning. The school van is late again and nobody told the parents. The queue at the county office has not moved. The shop where you buy lunch has run out of change, as it does every day. And you think: someone should build an app for this.
That thought is where a product can begin, and where many stop. Between "I see a problem" and "here is a product" there is a piece of work that has nothing to do with code. It is the work of narrowing: from a situation that annoys you, to a specific person with a specific cost, to one small job that software can do for them. This article walks through that work in order, using one example throughout.
Is every problem I see a product?
No. A problem becomes a product only when it passes three filters.
Someone owns it. You see the late school van. But who suffers the cost? Perhaps the parents, who worry. Perhaps the van owner, who loses clients. The owner of the problem is the person who would pay to make it go away, and it is often not the person who notices it.
It repeats. A problem that happens every day or every month can support a product. A one-off irritation cannot.
Money is already nearby. Somewhere close to the problem, money changes hands. Fees are collected, goods are sold, staff are paid. Software that sits near existing money can charge for itself. Software for a problem where no money moves has to create a budget from nothing.
Take the school van. Say you are a parent in Ruiru. You notice the vans are often late and the monthly fee is collected in a chaotic way. After a few conversations with van owners, you learn that lateness annoys parents, but what costs the owners real money is fees: parents pay late, pay part, or dispute what they owe. The owner of the costly problem is the van owner, it repeats monthly, and money is at the centre of it. That passes all three.
How do I write the problem down so that it can be built?
On one page, before you describe any solution. If you cannot fill this in, you are not ready to build, and you have found out for free.
The one-page problem brief
Who has the problem: (one kind of person, narrow enough that you could list ten)
What happens: (the problem in two sentences, in their words)
How often: (daily, weekly, monthly)
What they do about it today: (the workaround, step by step)
What it costs them: (shillings or hours per month)
Who pays for a solution: (the same person, or someone else)
The one result they want: (a single sentence, no features)
How I can reach them: (the group, place or person)
What I have not confirmed yet: (be honest)
Here it is filled in for the example. The figures are illustrative.
Who: Owners of one to three school vans in Ruiru and Juja, each carrying 15 to 40 children.
What happens: Parents pay the monthly fee late and in pieces, by M-Pesa to the owner's personal line. The owner cannot easily tell who has paid what.
How often: Every month, worst in the first two weeks.
Today: An exercise book, scrolling through M-Pesa messages, and phone calls to parents in the evening.
Cost: Several evenings a month, and perhaps KES 10,000 a month in fees never collected.
Who pays: The van owner.
One result: "I want to know at a glance who has paid this month and who has not."
Reach: The drivers at two schools I know, and their owners' WhatsApp group.
Not confirmed: Whether owners would pay a monthly fee for this.
Notice what the brief does not contain: the word "app", a list of screens, or live tracking of the van. Those were your first ideas. The brief has replaced them with what the customer said.
How do I work out which part software should do?
Map what happens today, one step at a time, and look for the step that costs the most. That step is your product. The rest can wait.
Sit with two or three customers and ask them to show you the whole process from beginning to end. Write each step on its own line.
| Step today | Who does it | Time or cost | Could software do it? |
|---|---|---|---|
| Tell parents the fee for the term | Owner, by WhatsApp | Small | Yes, but little gain |
| Receive payments on M-Pesa | Parents | None for the owner | Already works |
| Work out who paid and how much | Owner, from messages and the book | Hours each month, with errors | Yes, large gain |
| Remind parents who have not paid | Owner, by phone in the evening | Hours, and awkward | Yes, large gain |
| Handle disputes about balances | Owner and parent | Stressful, loses clients | Follows from step 3 |
| Tell parents when the van is late | Driver, sometimes | Small | Yes, but not what owners pay for |
Two steps stand out, and they are linked: knowing who has paid, and reminding those who have not. That is the product. A tool that matches incoming payments to each child and sends a polite reminder to parents with a balance.
The table has also told you what to leave out. Live tracking of the van, the thing you first noticed as a parent, is in the bottom row. It may come later. It is not what the paying customer loses sleep over.
Should my first version be an app?
Usually not. Choose the simplest form that delivers the result, and move up only when the form below it breaks.
- You, by hand. The owner forwards M-Pesa messages. You keep the list and send the reminders yourself on WhatsApp. This costs nothing to set up and can begin on Monday.
- A shared spreadsheet and message templates. Still manual, but faster, and the owner can look at it.
- A simple web tool. A page that opens from a link on any phone, with payments recorded automatically through a Paybill, with each child as the account number, and reminders sent by SMS or WhatsApp. No download needed.
- A mobile app. Worth it when customers use the product many times a day, need it to work without a connection, or need the phone's camera or location.
It is tempting to jump to level 4 because "app" is the word everyone knows. Each level up costs more to build and to maintain. Begin at level 1 with three to five paying customers. The manual weeks will show you things no plan would: that parents pay for two children in one payment, that some pay from a spouse's number, that owners want a statement to print for each parent. Every one of those is a requirement you would otherwise have discovered after paying for the build.
Move to level 3 when the manual work takes more hours than you have and customers are paying. For the van example, a web tool is likely the right long-term form. Owners check it a few times a week, and parents only need to receive a message.
What do I take to a developer?
Five things. With these, a competent developer or team can give you a realistic price and timeline. Without them, you will get a guess.
- The one-page problem brief.
- The process map, with the steps the product takes over clearly marked.
- The one job, in a sentence. "Match each M-Pesa payment to a child and remind parents who owe."
- The "not in version one" list. Van tracking, a parent app, attendance, expense tracking. Written down, so nobody adds them quietly.
- Your evidence. The number of customers you served by hand, what they paid, and the odd cases you met.
Then ask for three things in return. A fixed scope and price for version one. Ownership of the code, the domain and the accounts for hosting, payments and messaging, in your company's name if the company exists and otherwise in your own, never the developer's. And a first release within weeks, not months, to the customers you already have.
What mistakes turn a good problem into a bad product?
The same few, repeatedly.
- Building for the person who notices instead of the person who pays. The parent sees the late van. The owner pays for fee collection.
- Solving the visible symptom. What you see from outside is rarely the costly part.
- Describing the solution before the problem. "An app with tracking, payments, chat and ratings" is a guess at an answer to a question you have not finished asking.
- Adding features to feel safer. Each one adds cost and delay and makes the product harder to explain.
- Leaving payments for later. If the product is near money, how money gets in and is recorded is the core of it, not an extra.
- Skipping the manual stage because it feels unprofessional. Customers do not mind. They care that the result arrives.
A problem you see every day is a real advantage. You did not have to go looking for it, and you understand its setting. Do the narrowing work, and "someone should build an app for this" becomes a page that a builder can act on, and a first version small enough to finish.
Key takeaways
- Seeing a problem is not the same as owning it. Find the person who pays the cost and build for them.
- Put the problem on one page before you describe any solution.
- Map the current process step by step and pick the one step that costs the most.
- The first version is rarely an app. Start with the simplest form that delivers the result.
- Take a developer a brief, a process map and five waiting customers, not a list of features.
- Decide what is left out of version one and write that down too.
Frequently asked questions
- Do I need to learn to code to turn a problem into a product?
- No. Your job is to understand the problem and the customer better than anyone else and to describe the first version clearly. Someone else can write the code. A founder who can explain exactly what the product must do, and what it must not do, is easier to build for than one who can code but is unclear.
- How much does it cost to build the first version?
- It depends on three things, which are how many jobs the product does, whether it needs a mobile app or can run in a browser, and how many outside systems it connects to, such as payments and messaging. A first version that does one job costs a fraction of one that does six. That is the main reason to narrow the problem first.
- What if the problem I see is in an industry I do not work in?
- Then you see the symptom and not the cause. Spend time with people who do the work before you decide what to build. Sit with them for a day, ask them to show you how they handle it, and look for the step they complain about most. You may find the real problem is next to the one you noticed.
- Should I protect my idea before I talk to developers?
- Describing a problem to a developer puts you at little risk, because the value is in your customer knowledge and the work ahead. When you are ready to share detailed plans or hire someone, a simple written agreement covering confidentiality and who owns the code is sensible. Get a lawyer to draft it.
Sources
- The Lean Startup: How Today's Entrepreneurs Use Continuous Innovation to Create Radically Successful Businesses, Eric Ries (book, 2011). Checked on 19 November 2026.
- The Mom Test: How to Talk to Customers and Learn If Your Business Is a Good Idea When Everyone Is Lying to You, Rob Fitzpatrick (book, 2013). Checked on 19 November 2026.
