Stage 1 of 10 · from idea to first customer

Find a problem worth solving

Dig up real, painful problems in places where people already complain, before you fall in love with a solution.

A startup that solves a problem nobody cares about fails no matter how well it is built. This stage protects you from that before you spend a cent: you read what people complain about in public, in their own words, and you collect the problems that keep coming back.

Reddit threads, one-star reviews, job ads and freelance gigs are free research archives. Each method below shows where to look, which phrases to search and how to log what you find so patterns jump out.

What you should have at the end of this stage

A list of 3 to 5 specific problems, each backed by real quotes from real people, with a rough score for how frequent it is and how much it hurts.

The 10 methods in this stage

Method 1

Reddit Problem Research

Search subreddits for repeated complaints and wishes to find problems people already talk about, in their own words.

Beginner3 to 4 hours
Method 2

Amazon Review Mining

Read the middle-star reviews of best-selling products to find the gap buyers keep describing and nobody has fixed.

Beginner3 to 5 hours
Method 3

App Store Review Research

Read low-star reviews of competing apps to find missing features, update anger and users who are ready to switch.

Beginner3 to 4 hours
Method 4

Quora Question Analysis

Find questions people keep asking on Quora that get weak answers, then rewrite them as problem statements you can test.

Beginner2 to 3 hours
Method 5

Facebook Group Insights

Join niche Facebook groups, search past posts and ask one honest question to learn which problems members raise again and again.

Beginner1 to 2 weeks, about 30 minutes a day
Method 6

Problem Journal

Log every frustration you see for 2 to 4 weeks, then score each one on frequency, pain and willingness to pay.

Beginner10 minutes a day for 2 to 4 weeks, plus 2 hours to score
Method 7

Industry Forum Research

Read the forums, Stack Exchange sites and Slack or Discord groups of one profession to find the problems they solve by hand.

Intermediate4 to 6 hours over a week
Method 8

Job Board Signal Mining

Read job ads for repetitive manual tasks: when companies hire people to do them, they already spend money on the problem.

Intermediate3 to 4 hours
Method 9

Freelance Marketplace Research

Study what people repeatedly pay freelancers for on Upwork and Fiverr, at what price, and which tasks you could productize.

Intermediate3 to 5 hours
Method 10

Competitor Review Analysis

Read low-star reviews of market leaders on G2, Capterra and Trustpilot and map why customers switch away from them.

Intermediate4 to 6 hours

Go deeper