Most weak ideas start as a sentence: “an app for X.” Most usable ideas start as a job someone already pays to get done, badly.
You do not need a once-in-a-lifetime insight. You need a problem that shows up every week, a person who feels it, and a way they already spend money on it — an agency, a spreadsheet, an accountant, a Zalo thread, overtime.
#Look at work that already happens
Talk to people who run a company, not people who “have an idea.” Ask what they did last Tuesday that they hated. Ask what they forwarded to someone else. Ask what they waited on.
Patterns that often turn into companies:
- A step that only one person in the company can do, so everything queues
- A file that lives in chat, then gets retyped into software
- A delay between finishing the work and sending the bill
- Two people keeping two copies of the same truth
If nobody is paying for a workaround today, they probably will not pay you tomorrow.
#Do not start from the tool
“AI,” “platform,” and “marketplace” are not ideas. They are how you might later build. The idea is who, what pain, and what they do instead.
A founder who already issues invoices, chases payment, and sits with an accountant has seen more real problems than a pitch deck. Write those down before you invent a category.
#Steal the shape, not the clone
It is fine if a similar product exists abroad. The useful question is whether Vietnamese companies hit the same wall — tax invoices, MST, the accountant’s queue, cash stuck waiting on customers. A copy of a US workflow often fails here because the painful step is different.
#Next
- Evaluate ideas — which ones are worth building
- Wedge market — the first customers you can actually win