Look for workarounds, not trend words
Profitable problems usually announce themselves through behavior: a spreadsheet maintained by three people, a daily copy-and-paste ritual, a queue nobody owns, a customer question answered differently by every employee, or a report assembled after the decision is already due.
These workarounds are evidence that the problem is real enough to consume time. Your task is to learn whether it is painful, frequent, costly, and attached to a buyer with authority.
Use the five-signal filter
Score the problem before designing the solution. Strong opportunities combine frequency, consequence, budget, reachable buyers, and a change in technology or behavior that makes a better approach possible now.
- Frequency: it recurs often enough to justify a system.
- Consequence: delay or error has a visible cost.
- Budget: someone already spends money, labor, or attention on it.
- Reach: you can identify and contact the person who owns the outcome.
- Timing: a new capability or constraint changes the economics.
Interview the past, not the imagination
Ask for the last time the problem occurred. What triggered it? Who became involved? What did they try? How long did it take? What went wrong? What happened next? Specific history produces better evidence than asking whether somebody would use an imagined product.
Listen for repeated nouns, systems, approvals, and moments of delay. Do not pitch during the interview. Your goal is to understand the workflow well enough to describe it back accurately.
A complaint is interesting. A repeated workaround with an owner and a cost is a market signal.
Keep an evidence ledger
Record the claim, source, date, buyer type, current workaround, estimated frequency, consequence, and confidence. Mark what you observed directly, what the buyer reported, and what you inferred. This prevents enthusiasm from turning a single conversation into a fictional market.
After five to ten conversations, patterns should become sharper. If every workflow is different and nobody owns the outcome, the market may be too fragmented for a repeatable offer.
Know when to reject the idea
Walk away when the pain is ceremonial, the buyer cannot be reached, the consequence is trivial, the data cannot be used responsibly, or the solution depends on accuracy you cannot deliver. Also reject problems that disappear when a buyer uses an existing feature they already pay for.
The discipline to abandon weak problems is an edge. It preserves time for the few situations where urgency, economics, and new leverage genuinely line up.
Content history See what changed
Added the five-signal problem filter, interview prompts, evidence ledger, and explicit rejection criteria.
First published.