From Customer Questions to a Business Idea: Keep a Problem Notebook

Tom — Singapore Digital Business & Systems Consultant

Tom

Digital Business & Systems Consultant · Singapore

7 min read Article date First published
From Customer Questions to a Business Idea: Keep a Problem Notebook

To find business ideas from customer problems, write down the problems you keep seeing: the situation, who has it, how often, what people do about it now and what that costs them. Then test only the few entries that keep coming back with real evidence. Most entries will never become a business, and that is fine. The notebook trains you to notice patterns, not to treat every passing complaint as an opportunity.

Key takeaways

  • A problem notebook records patterns, not people: write roles and situations, never names or private details.
  • The current workaround and its cost in effort tell you more than the complaint itself.
  • Test an entry only when it repeats across different people and someone already spends effort working around it.
  • Most entries will be parked; that is the notebook doing its job. Even a proven problem often needs a form before an app.

Where useful problems show up

Repeated friction is already around you:

  • Questions customers repeat at work. The same question five times in one shift says something about how information reaches people.
  • Family and friends' small businesses. The admin they grumble about after closing.
  • Community and parent groups. Requests that return to the group chat every week.
  • Your own routines. Anything you retype, chase or work around.

Restaurant, retail and event jobs teach far more than customer service. Observe patterns there, but do not take employer documents or data, and do not approach your employer's customers for your own business.

Write roles, not names. Never record customers' names, numbers, photos or chat screenshots, and never copy an employer's SOPs, sales figures, prices or client lists. A pattern needs no identity.

Record the situation and the workaround

Each entry starts with what happened and where. Then note the workaround: what people do today to cope. A workaround shows the problem matters enough for someone to act, and it is what any idea of yours would have to beat. A problem with no workaround is often one nobody minds much.

The problem notebook: five columns

Copy these columns into a notebook, notes app or spreadsheet. The entries come from the fictional example below.

ColumnWhat to writeExample entry (fictional)
SituationWhere and when it happened, described by roleWeekday evening; a customer rescheduling a servicing visit by chat
EvidenceHow often, and what kind of person, not whoFourth time this week; mostly working adults at home on weekends
Current workaroundWhat people do now to copeRetyping the change into a spreadsheet and sending a reminder by hand
Cost in effortTime, rework, mistakes, waiting or moneyAbout 30 minutes most evenings; one double booking this month
Next questionOne neutral thing to ask or observe nextDo other small service businesses handle changes the same way?

At each review, add a status beside the entry: watching, parked or testing.

Ask what the problem costs in effort

This column separates a mood from a problem. A complaint about queue music is a mood; retyping the same order details three times a day is a cost. Ask how often it happens, how long it takes and what goes wrong when it is missed. If nobody loses time or money over it, note it and move on.

Test whether others share it

Keep your next question neutral. "Would you use an app for this?" invites a polite yes. Ask what already happened:

  • "When did this last happen?"
  • "What did you do about it?"
  • "How long did that take?"
  • "Have you tried anything else?"

My rule of thumb: move an entry from watching to testing only when you can tick at least four of these five boxes.

  • I have seen it at least three times, from more than one person or place.
  • People already spend time, effort or money on a workaround.
  • I can describe who has the problem in one sentence.
  • At least two people described the last time it happened without me suggesting a solution.
  • I could try a small manual fix myself, without customer data or employer material.

A passing entry is still a hypothesis. The next step is a small, low-cost test, not a website or app.

Turn a promising entry into a small experiment with a budget cap and a stop rule.

Read: How to Test a Business Idea Before Paying for a Website →

Most entries will not become businesses, and that is fine

Some problems happen once. Some are already solved well. Some are real but too small, or need skills or licences you do not have. Parking them is not wasted effort; each one sharpens your eye. Review the notebook monthly and decide what to keep watching, park or test. Exploring one industry over a few months? A monthly learning journal pairs well with it.

A fictional example: six weeks of entries

(Fictional example) Siew Lin keeps the books for her family's two-person aircon servicing business. For six weeks she keeps a problem notebook, writing roles and never customer names.

Most entries go nowhere. Neighbours asking the block chat for a handyman come up often, but a recommendation settles each request in a minute, so she parks it. A friend's complaint about delivery fees never recurs.

One entry keeps returning: customers rescheduling by chat, which she retypes into a spreadsheet each evening. She asks two other small service owners she knows how they handle changes; both describe the same retyping before she suggests anything.

She tests the simplest fix first: an online form feeding her own spreadsheet, which takes four weekends to set up and adjust. Retyping drops, but reminders are still sent by hand, and one older customer prefers to call. Whether other owners would pay for this setup, or whether a custom booking tool is ever justified, is still unknown. Her next question is what the two owners already spend on scheduling.

When a spreadsheet or form is enough, and when a tool makes sense

Even a proven problem rarely needs software first. Try a shared spreadsheet, an online form, a checklist or a template message. If that works, you may have a useful service with no custom software at all. Consider a custom tool only when:

  • The simple fix has run for weeks and the workflow has settled.
  • Several people need the same records at the same time.
  • You can count the hours or errors the manual steps still cost.
  • Ready-made tools you tried do not fit how the work is done.

A repeated internal workflow, such as jobs and changes passing through WhatsApp and spreadsheets, is what operations digitalisation addresses. A booking tool, client portal or dashboard is web app development, which I scope and quote individually. Development is a real cost, so it only makes sense once the problem is proven and the simpler fix has run out of road.

Your next step

Set up the five columns today and add entries for two weeks before judging any of them. If an entry passes your rule and a manual fix is now the bottleneck, message me a short summary, without customer names. Sometimes my honest answer will be to keep the spreadsheet.

Frequently Asked Questions

How do I know a problem is worth solving?

Look for a pattern with a cost: the problem repeats across different people, they already spend effort or money working around it, and they can describe the last time it happened. One loud complaint is not enough.

Should I build an app for every repeated problem?

No. Many repeated problems are handled well by a form, a shared spreadsheet or a checklist. Consider a custom tool only after a simpler fix has been tried and the workflow is proven, frequent and shared by several people.

What should I never write in a problem notebook?

Customers' names, phone numbers, photos, chat screenshots or other personal details, and any employer's confidential information such as SOPs, sales figures or client lists. Describe roles and patterns instead.

Can I use problems I notice at my part-time job?

You can learn from what you observe, such as the kinds of questions customers ask. Do not take employer documents or data, and do not approach your employer's customers for your own business.

What if someone already solves the problem?

That is useful evidence. Note what the existing solution misses and who it leaves out. Sometimes the honest conclusion is that the problem is already well handled.

Has one notebook entry passed your test, and is a manual workaround now the bottleneck? Message me a short summary, with no customer names, and we can talk through whether a form, a spreadsheet or a scoped web app fits. It is a conversation, with no commitment.

No obligation. No sales pitch. Just an honest conversation.

お問い合わせ