What to Research First — A Framework for Startups and Independent Builders

Filed under: Research Strategy · 7 min read

You've decided to do market research. Now the real question: what do you research first?

Most people freeze here. The list of things they don't know is long. The list of things they think they know is longer. Without a clear first target, research becomes a way to avoid making decisions — not a way to inform them.

The Decision Matrix: Impact vs. Uncertainty

Every research topic can be scored on two axes:

• Impact — how much does this question shape your decision? If you knew the answer, would it change what you build, who you target, or how you position?
• Uncertainty — how confident are you right now? Low confidence = high uncertainty = high research value.

Your first research topic should always be high impact + high uncertainty. These are the questions that are blocking a major decision.

The Three Categories of Research

1. Market Validation Research
Does this market exist? Is there demand? Who specifically has the problem? What do they pay for today?
Example: "Is there a real market for AI coding agents in mid-market enterprises?"

2. Competitive Landscape Research
Who else is solving this? What's working for them? What do customers complain about? Where are the gaps?
Example: "What's the pricing landscape for project management tools in the $10-50k ACV range?"

3. Opportunity Sizing Research
How big is this market? Is it growing? What segment is most underserved?
Example: "What's the total addressable market for AI-powered legal document review?"

Start with Market Validation. If there's no market, competitive and opportunity research are academic. Validate first.

Questions That Make Bad First Research Topics

"I want to research AI trends."

This is too broad. "AI trends" spans everything from healthcare to finance to retail. Break it down: "What are the top 3 pain points for engineering leaders evaluating AI coding tools in companies with 50-200 engineers?" — that's a researchable question.

"I want to know everything about my competitor."

Know one specific thing: what do their customers complain about most? That's your entry point. You don't need a full competitive analysis — you need a wedge.

The Test: Would Knowing Change What You Do?

Before you finalize your research topic, ask: if the brief comes back and confirms X, what changes? If the answer is "I'd change my pricing," research pricing. If the answer is "I'd give up entirely," that's a high-impact question worth paying for. If the answer is "I'd keep building the same thing," the research has no decision impact — skip it.

The goal of research is not to feel informed. It's to make a better decision than you would have without it.