Not every problem is worth solving. Some are too small to matter. Some are too big to tackle. Some look like problems but turn out to be preferences.
The best problems share a specific signature: repetition.
When someone does the same manual work every week — gathering the same type of information from the same sources for the same purpose — that’s not just a task. That’s a product waiting to exist.
The Repetition Signal
One-time problems are consulting projects. Recurring problems are businesses.
If someone needs to pull data once for a specific decision, they’ll do it themselves or hire someone. The pain isn’t severe enough to justify a subscription.
But if they need to pull the same data every week, month, or quarter — if the workflow never ends — the calculus changes. Three hours of manual work times fifty-two weeks is 156 hours a year. That’s almost a month of full-time work, every year, forever.
Now there’s a budget.
Repetition signals that the problem is structural, not situational. It’s not going away. The person experiencing it will keep experiencing it until something changes. And if you can be that change, you can charge for it indefinitely.
Where to Look
The best product opportunities are hiding in plain sight. You just have to know what to look for.
Watch for scheduled tasks. When someone blocks the same time every week to “update the spreadsheet” or “pull the reports,” that’s a flag. Calendar appointments for data gathering are symptoms of unsolved problems.
Listen for complaints about sources. “The county website is so slow.” “I have to check six different sites.” “The format changes every time.” These aren’t just venting — they’re feature requests disguised as frustration.
Notice the workarounds. Bookmarked URLs. Copy-paste routines. Spreadsheets with manual formulas. Chrome extensions that sort of help. When people build elaborate systems to work around a problem, they’ve already validated that the problem matters.
Ask about the first hour of the week. Most professionals have some version of “Monday morning prep” — time spent gathering information before the real work begins. That prep time is often where the best product opportunities live.
Follow the junior employees. Repetitive data tasks tend to roll downhill. If a company has someone whose job includes regular manual research, that’s a role that could partially or fully be replaced by a product.
The Validation Shortcut
Here’s the fastest way to validate a product idea: find people already doing the work manually and ask if they’d pay to stop.
Not “would you be interested in a product that…” — that’s too abstract. People say yes to hypotheticals.
Instead: “You mentioned you spend three hours every week pulling foreclosure filings. If I delivered that data to your inbox every Monday for $150/month, would you sign up?”
Specific. Concrete. Requires a real answer.
If multiple people say yes to the same specific offer, you’ve found something worth building. If they hesitate, ask why. Their objections will tell you what the product actually needs to be.
The goal isn’t to survey the market. It’s to find ten people who will pay before you build anything.
Good Problems vs. Bad Problems
Not all repetitive tasks make good products. Some filters:
Is the data source accessible? Government websites, public APIs, and open databases are fair game. Proprietary systems, login-gated platforms, and data behind legal restrictions are usually not worth the complexity.
Is the workflow standardized enough? If every customer needs completely different fields, formats, and sources, you’re not building a product — you’re building an agency. The best problems have common patterns across multiple potential customers.
Is the frequency high enough? Weekly or monthly tasks create subscription businesses. Annual tasks don’t — the pain isn’t acute enough to justify ongoing payment.
Is the audience reachable? You can build the perfect product for a market you can’t access. Niche professionals in identifiable communities (forums, associations, LinkedIn groups) are easier to reach than scattered individuals.
Is there budget authority? The person experiencing the pain needs to be able to pay for the solution — or be able to convince someone who can. $150/month is an easy expense report. $1,500/month requires approval.
The Interview Framework
When you’re exploring a potential problem, here’s what to ask:
“Walk me through how you currently do this.” Get the full picture. Every click, every source, every step. The details reveal where the real friction is.
“How often do you do this?” Establishes the repetition. Weekly is better than monthly. Monthly is better than quarterly.
“How long does it take?” Quantifies the pain. Multiply by frequency to get annual time spent.
“What happens if you don’t do it?” Reveals the stakes. If the answer is “nothing much,” the problem isn’t painful enough. If the answer is “I miss deals” or “I get fined,” you’ve found something real.
“What have you tried to make this easier?” Uncovers existing solutions and their shortcomings. Also validates that the person has already invested effort in solving this — a strong buying signal.
“If this showed up in your inbox every Monday, ready to use, what would that be worth to you?” The money question. Don’t skip it. The answer tells you whether this is a real problem or just a mild annoyance.
From Problem to Product
Once you’ve found a problem worth solving, the path forward is straightforward:
Solve it manually first. Pull the data yourself, by hand, for a few customers. This validates that your solution actually works and surfaces all the edge cases you’d miss by theorizing.
Charge from day one. If people won’t pay for the manual version, they won’t pay for the automated version. Price validates value.
Then automate. Once you’ve proven the model works, build the systems that let you deliver without the labor. Your manual process becomes your spec for what to build.
The best products don’t start with ideas. They start with problems — specific, recurring, painful problems that people are already trying to solve.
Find the repetition. That’s where the opportunity lives.
Headwater AI builds data products that eliminate repetitive manual work. We find the workflows that waste hours every week and turn them into automated feeds.