After analyzing hundreds of market research queries, we've seen the same patterns that lead to useless results. Here's what to avoid.
1. Confirmation bias disguised as research
The most common mistake: you've already decided your idea is great, and you're looking for data to confirm it. This isn't research — it's cherry-picking.
Fix: Ask proofnest explicitly: "What are the strongest arguments against this idea?" The self-challenge layer does this automatically, but framing it up front forces even deeper scrutiny.
2. Researching a solution instead of a problem
"I want to build an AI chatbot for restaurants" isn't a problem statement — it's a solution. The real question is: what problem do restaurant owners have that a chatbot could solve?
Fix: Start with the pain. "Restaurant owners spend 2 hours daily answering the same phone questions about hours, reservations, and menu items." Now you're researching demand for a real problem.
3. Ignoring existing solutions
"Nobody has built this" is almost never true. More likely: you haven't found them yet. And if truly nobody has built it, that might be a red flag — not a green one.
Fix: proofnest scans dozens of sources to find competitors you didn't know about. But also ask yourself: if the problem is real and the market is big, why hasn't someone solved it? The answer tells you a lot.
4. TAM fantasy
"The global CRM market is $80 billion, so if we just capture 0.1%..." This is not market sizing. This is daydreaming with decimals.
Fix: Start from the bottom up. How many specific people have this specific problem? How many of them would pay? At what price? proofnest's TAM/SAM/SOM analysis does this with real data, not top-down fantasy.
5. Skipping the "why now?" question
Every idea needs a timing catalyst. Why is this possible now and wasn't 2 years ago? New technology? Regulatory change? Cultural shift? Platform growth?
Fix: Include timing context in your prompt. proofnest's research will pick up on trends and growth signals, but framing the "why now" makes the analysis more focused.
6. Falling in love with a niche of one
You found 3 Reddit posts asking for your exact product. That's not validation — that's a niche of one (or three). Real demand shows up across multiple channels, communities, and data points.
Fix: proofnest cross-references signals across Reddit, ProductHunt, Google Trends, SEO data, app stores, and more. If demand only shows up in one place, the report will flag it.
7. Not defining "kill triggers" before starting
If you don't decide upfront what would make you walk away, you'll rationalize any finding. "The market is crowded, but our UX will be better" is the graveyard of startups.
Fix: Before you run research, write down 3 things that would kill the idea. Then check the report against them honestly. proofnest includes kill triggers in every action plan — take them seriously.
The meta-lesson
Good market research isn't about finding reasons to proceed. It's about finding reasons to stop — and not finding them. The ideas that survive honest scrutiny are the ones worth building.