Market Research
Overview
Market research answers the questions your gut cannot: How big is this really? Who's already here? What's coming? What do customers actually look like? This playbook gives you a repeatable system using mostly free tools. Run it before launching, and refresh quarterly.
Step 1: Define Your Research Questions First
Never open a browser without knowing what you're looking for. Write 3-5 specific questions. Everything you find must map back to one. Examples:
- How large is the addressable market for [X]?
- Who are the top 5 competitors and what are their actual weaknesses?
- What macro or industry trends will shape this space in the next 12-24 months?
- What does a typical customer in this market look like demographically and psychographically?
- What price points do existing solutions charge, and where are the gaps?
Step 2: Market Sizing (TAM → SAM → SOM)
You do not need exact numbers. Order-of-magnitude is enough to make decisions.
TAM (Total Addressable Market): Everyone on earth who theoretically has this problem or need. Often found in industry reports (Google "[industry] market size report").
SAM (Serviceable Addressable Market): The slice of TAM you could realistically reach given your language, geography, channel, and positioning. Apply realistic filters: "Of the global TAM, how many are in English-speaking markets? How many use the platform/channel I can reach?"
SOM (Serviceable Obtainable Market): What you can realistically capture in Year 1. Typically 1-5% of SAM for a new entrant.
When no report exists — estimate from first principles:
- Find a related population stat (e.g. "There are 60M freelancers in the US").
- Apply a conversion funnel: "20% use project management tools → 12M. Of those, 10% have the specific pain I'm solving → 1.2M."
- Multiply by realistic ARPU: "If each pays $15/month → $216M TAM."
- Sanity-check against competitor revenue if any are public.
Decision rule: If your realistic SOM < $500K annual revenue potential, the market is too small to sustain a solo business at a healthy margin. Consider a broader niche or higher price point.
Step 3: Free Data Source Toolkit
Market size & industry data:
- Google "industry name market size" — even headline numbers from reports are useful
- Statista (free snippets give key stats)
- IBISWorld, Grand View Research — free summaries
- SEC filings of public companies in the space (revenue, growth, strategy)
- Bureau of Labor Statistics — workforce and industry employment data
Competitive & product intelligence:
- Competitors' websites — pricing pages, feature lists, careers pages (hiring reveals strategy), blog topics (reveals what they think matters)
- G2, Capterra, Trustpilot, App Store — read 20+ reviews minimum. Categorize complaints.
- SimilarWeb (free tier) — traffic volume and source breakdown
- Crunchbase (free tier) — funding history, investors, founding dates
- LinkedIn company pages — headcount trends (growing = healthy signal)
Trend & signal intelligence:
- Google Trends — search interest over time
- Reddit — search within relevant subreddits for recurring themes
- Product Hunt — what's gaining traction in adjacent spaces
- Y Combinator batch announcements — what's getting funded signals what's coming
- Twitter/X — real-time industry conversation
- Industry newsletters — find 3-5 relevant ones and read for 2 weeks before deciding
Step 4: Competitor Landscape Mapping
Map your top 5-7 competitors (or adjacent players if direct competitors are few). For each, build a profile:
NAME: [Company name]
URL: [link]
FOUNDED: [year]
FUNDING/REV: [what's publicly known]
TARGET: [who they serve — be specific]
VALUE PROP: [one sentence: why customers choose them]
PRICING: [model + price points]
TOP FEATURES: [3-5 things they do well]
STRENGTHS: [from their marketing + positive reviews]
WEAKNESSES: [from negative reviews and feature gaps]
TRAFFIC: [SimilarWeb estimate]After profiling all competitors, synthesize into three lists:
- Table stakes — things ALL competitors do. You must match these or customers won't consider you.
- Universal gaps — things ALL competitors do poorly or skip entirely. This is your opportunity.
- Underserved segments — customer types that no competitor targets well. This is your niche entry point.
Step 5: Trend Analysis
Identify 3-5 trends relevant to your market. For each, assess:
- What is the trend? (be specific, not "AI is big")
- Evidence: 2-3 concrete data points or examples.
- Maturity: Emerging (< 20% adoption) / Growing (20-50%) / Maturing (50%+) / Declining.
- Impact on your opportunity: Tailwind (helps you) / Headwind (hurts you) / Neutral.
- Action: What should you do because of this trend? (Build for it / Watch it / Ignore it / Avoid it)
Step 6: Customer Persona Construction
Build 2-3 personas. These are not marketing fluff — they are decision-making tools. Every product, pricing, and messaging decision should be testable against "would Persona X care about this?"
PERSONA NAME: [fictional but grounded in data]
ROLE & INDUSTRY: [specific]
SENIORITY: [junior / mid / senior / founder]
COMPANY SIZE: [solo / 2-10 / 11-50 / 50+] (if B2B)
LOCATION: [region or work style: remote/hybrid/office]
ANNUAL INCOME: [range — relevant for pricing decisions]
DAILY REALITY:
- What does their typical workday look like around [your problem area]?
- What tools do they already use in this workflow?
PAIN POINTS: [top 3, ranked by severity — pull from reviews and interviews]
GOALS: [what they're trying to achieve professionally]
BUYING BEHAVIOR: [how they discover tools, decision timeline, budget authority]
CHANNELS: [where they spend time online — critical for marketing later]
QUOTE: [a real or realistic quote capturing their mindset — pull from forum posts or interviews]Step 7: Compile and Act
Produce a single Market Research document containing:
- Market size (TAM/SAM/SOM with sources)
- Top competitor profiles + synthesis (table stakes, gaps, underserved segments)
- 3-5 trends with impact assessment
- 2-3 customer personas
- Recommended next actions — numbered, specific, prioritized
Schedule a quarterly refresh. Markets move. What's true today may not be in 90 days.