How to Find Problems Worth Solving: A Founder's Playbook

Most startup advice says "solve a real problem." That's incomplete. Lots of real problems aren't worth solving commercially. The problem needs to be real, painful, frequent, and underserved — all four at once.

Miss one and you end up with a product that works but doesn't sell, or sells but doesn't retain, or retains but can't grow. Finding problems worth solving is the most important skill a founder can develop.

Here's how to develop it systematically.

The Problem Worth Solving Framework

Not all problems are created equal. Before investing time in research, validation, or building, run every problem through this four-part filter.

Frequency: Does it happen often enough?

A problem that someone encounters once a year is not a business. A problem they encounter every week is. The more frequently someone runs into a problem, the more likely they are to seek a solution, pay for it, and stick with it.

Questions to ask: How often does this problem occur in a typical workflow? Daily, weekly, monthly, rarely? Do people actively seek solutions or passively accept it?

High frequency problems drive daily active usage and strong retention. Low frequency problems produce one-time buyers and forgotten apps.

Severity: Does it hurt enough to pay?

Mild annoyances don't generate revenue. Real pain does. Severity determines willingness to pay more than any other factor.

Look for problems that cost people something measurable: time, money, missed opportunities, or emotional stress. "This invoicing software is slightly annoying" won't make someone switch. "This invoicing software causes me to miss payments and look unprofessional to clients" absolutely will.

Severity signals on Reddit: long posts with specific details, emotional language ("nightmare," "embarrassing," "costing me clients"), multiple threads on the same problem, high upvote counts on complaint posts.

Market size: Are enough people affected?

A severe, frequent problem affecting only 50 people worldwide isn't a business. You need enough people to build a sustainable revenue base.

The minimum viable market for a bootstrapped SaaS: around 10,000 people who would pay $10-30/month. That's $1.2M-$3.6M in potential annual revenue if you can reach them all — a realistic target for a small team.

Don't fixate on total addressable market (TAM) the way VCs do. Focus on serviceable addressable market (SAM): the specific slice you can realistically reach and serve.

Underservedness: Are existing solutions inadequate?

If 10 excellent tools already solve this problem well, you're not finding a problem worth solving — you're finding a market that's already served. The question isn't whether solutions exist, but whether they're good enough.

The right question is: "Would someone who uses the best existing solution describe it as 'good enough' or as 'I use it because there's nothing better'?" The latter is your opportunity.

Where to Find Problems Worth Solving

The best sources for problem discovery are the places where people complain honestly.

Reddit communities

Reddit is the largest collection of unfiltered customer frustrations on the internet. For any professional or consumer audience, there are subreddits where members share their genuine problems without any filter.

The key is knowing how to search. Don't look for "pain points." Look for the language of frustration:

• "I've been struggling with..."
• "Anyone else find it impossible to..."
• "I hate that [tool] doesn't..."
• "Is there a better way to..."
• "I can't believe there's no tool that..."

Read the subreddit analytics guide to learn how to systematically mine these communities.

G2 and Capterra reviews

People write reviews when they're either delighted or frustrated. Focus on the frustrated ones. G2 and Capterra reviews, especially 1-star and 2-star entries, contain structured descriptions of problems that existing tools fail to solve.

Search for the leading products in any software category. Read the worst reviews. Group the complaints by theme. The themes that appear most often are the problems that aren't being solved.

Twitter/X searches

Search for "[product category] frustrating" or "I hate [specific tool]" on Twitter. Real-time complaints from real users. Less structured than Reddit but can surface emerging problems quickly.

Job postings

Companies hire people to solve problems they can't automate or tool away. A surge in job postings for a specific role often signals that no software tool is doing that job well yet. That's an opportunity.

Your own professional life

Problems you personally experience are worth investigating. You have deep context, you know the severity firsthand, and you can validate the solution with your own use. The caveat: confirm your experience is common. Many founders build for a market of one.

Scoring Problems Systematically

Once you've identified candidate problems, score them before investing time in deeper research.

Use a simple scoring rubric (1-5 for each dimension):

| Dimension | 1 (Weak) | 3 (Moderate) | 5 (Strong) | |-----------|----------|--------------|------------| | Frequency | Yearly | Monthly | Daily/Weekly | | Severity | Mild annoyance | Real friction | Costs money/time | | Market size | Under 1K potential users | 10K-100K | 100K+ | | Underservedness | Good solutions exist | Mediocre solutions | No good solutions | | Willingness to pay | Nobody pays | Some pay reluctantly | People actively pay |

Problems scoring 20+ out of 25 are worth deep research. Problems below 15 need a rethink.

This isn't a perfect system. It's a filter to focus your attention. Spend your research time on the high-scorers.

The Validation Test

Scoring a problem against a framework is hypothetical. Validation makes it real.

The Reddit post test. Write a post in the relevant subreddit describing the problem and asking if people experience it. Don't mention your solution. Just describe the problem accurately and ask "is this something you deal with?"

If 20+ people comment saying "yes, constantly" — you have a real problem. If nobody responds, either the problem isn't widespread, or you're in the wrong community.

The DM test. Find 10 people who complained about the problem publicly. DM them a simple note asking about their experience. Offer nothing. Just ask for 15 minutes of their time to understand the problem better.

Five conversations will reveal more than 500 posts. You'll hear the real severity, the solutions they've already tried, and — critically — whether they'd pay. Read the full approach in our customer discovery guide.

The landing page test. Build a one-page description of a hypothetical solution. No product required. Measure email signups. 50 signups in a week from strangers is a strong signal. Fewer means the problem or the positioning isn't connecting.

Problems Worth Avoiding

Knowing what not to build is as valuable as knowing what to build.

Vitamin problems. A vitamin makes you healthier over time. A painkiller stops immediate pain. Vitamins are nice to have. Painkillers are essential. Build painkillers. Vitamins have terrible retention because users don't feel the impact of not taking them.

One-time problems. If someone encounters the problem once and then never again, they'll solve it manually and never need your product again. Build for recurring problems.

Awareness problems. Some problems people don't know they have until someone shows them. These are real, but they require expensive education to market. Unless you have significant capital, start with problems people are already actively searching for solutions to.

Solved problems. Sometimes a problem looks underserved but isn't. The existing solution is just hard to find. Before assuming the market gap exists, do thorough competitive research using the market gap framework.

The Compounding Value of Problem-Finding Skill

Founders who build problem-finding into a systematic practice are never short of ideas.

They're reading Reddit regularly. They're monitoring competitor review sites. They're talking to people in industries they're curious about. They're building a backlog of scored, partially validated opportunities that they can pursue whenever they're ready.

The best founders aren't lucky. They're more systematic than everyone else about finding the problems that matter.

Stop waiting for inspiration. Start looking for frustration.

Keep Reading

• How to Find a Niche Market in 2026 — Narrow your focus to a specific audience
• The Market Gap Framework — Score opportunities before committing
• Why Most SaaS Products Fail — The patterns that kill products, all rooted in the wrong problem
• How to Find Validated SaaS Ideas on Reddit — Turn problems into concrete business ideas