Decide What You Actually Need to Learn
Most market research wanders because there is no question driving it. Before you open a single tab, write down the three things that would change your decision. Usually they are: are enough people looking for a solution, who already sells to them, and is there a slice those incumbents ignore. Everything else is trivia.
Framing it as a go or no-go decision keeps you honest. You are not trying to prove the idea is good. You are trying to find the fastest evidence that it is bad, because if you cannot kill it cheaply, that is itself a positive signal.
- Demand: are people actively searching for or complaining about this problem?
- Competition: who solves it today, and how happy are their customers?
- Wedge: is there a specific underserved group you could win first?
Size Demand Without a Spreadsheet Fantasy
The classic mistake is calculating a giant total addressable market by taking a huge population and multiplying by a price. That number is fiction and it convinces no one, including you. What you want instead is grounded evidence that people are actively looking for a fix right now.
Search behavior is the cleanest free signal. If hundreds or thousands of people search a problem phrase every month, demand is real and measurable. If almost nobody searches it, you may be inventing a problem. Pair search data with the volume of complaints you can find, and you get a directional read that is far more useful than a top-down market-size slide.
- Use a keyword tool to check monthly search volume for the problem and for any existing solutions.
- Count active discussions: how many recent threads, questions, and posts exist about this pain?
- Look at whether people are paying for adjacent tools, which proves budget exists in the category.
- Be suspicious of zero competition and zero search volume together. That usually means no market, not an untapped one.
Study Competitors Through Their Customers
Do not study competitors by reading their homepages. Their marketing tells you what they wish were true. Study them through their customers instead, because that is where the unmet needs leak out. Review sites, support forums, and cancellation complaints reveal exactly where existing products fall short.
The pattern you are hunting for is a consistent complaint that nobody has fixed. When ten different customers of the market leader say the same thing is missing or broken, you have found a real opening. That repeated gripe is more valuable than any feature comparison chart, because it is a gap the incumbent has chosen not to close.
- Read the three-star reviews first. They come from engaged users who want the product to be better.
- Search for 'alternative to [competitor]' to see what frustrated customers are shopping for.
- Note pricing and who it locks out. Often the gap is a cheaper, simpler option for people the leader ignores.
- Make a short list of the complaints that repeat. Repetition is the signal.
Find Your Wedge, Not the Whole Market
You will not beat an established player by being a slightly better version of them for everyone. You win by owning a narrow beachhead first: one specific group whose needs are poorly served and who can be reached cheaply. From that foothold you expand, but you start small on purpose.
The right wedge usually combines a clear underserved segment with a complaint the incumbents ignore. If review-reading turned up a repeated gripe from a specific type of user, that intersection is your wedge. It is narrow enough to win and real enough to grow from.
Talk to Five People to Pressure-Test It
Desk research gets you most of the way, but it cannot confirm that people will actually act. Five honest conversations with people in your target segment will either reinforce what you found or quietly demolish it. Both outcomes are worth far more than another day of reading.
Ask about their current behavior, not about your idea. How do they handle this problem today, what does that cost them, what have they tried and abandoned. If they describe a painful, expensive workaround in vivid detail, your research is confirmed. If they shrug, the market you saw online may not be as hungry as it looked.
- Recruit from the same forums and review threads where you found the complaints.
- Ask 'walk me through the last time this happened' instead of 'would you use this?'
- Listen for emotional language and real spending. Both signal a problem worth solving.
- If five people in a row are lukewarm, treat that as a no-go even if the search data looked promising.