■ Market Research Deep Dive Handbook
1) What is Market Research (Really)?
Market research is a structured way to reduce uncertainty about your market: customers, problems, competitors,
pricing, and demand. It collects and analyzes evidence so product, marketing, and leadership teams can make
decisions with confidence instead of guesswork.
● Inputs: questions, assumptions, early concepts, hypotheses.
● Activities: gathering data (primary & secondary), analyzing it (quant + qual), and synthesizing insights.
● Outputs: customer insights, personas/JTBD, opportunity size (TAM/SAM/SOM), pricing guidance, positioning,
and concrete recommendations.
How it differs from analytics & UX research
Web/Product analytics show what current users do in your product.
UX research focuses on usability and experience of a solution.
Market research zooms out: size, preferences, willingness to pay, competitors, and unmet needs—even among
non■users.
2) Why It Matters for Product & Growth
● De-risks launches (build what people value, not just what’s feasible).
● Validates positioning and pricing before big spend.
● Reveals which customer segments are worth pursuing now vs later.
● Turns fuzzy opinions into prioritized, testable roadmaps.
3) Types of Market Research — When to Use What
● Primary vs Secondary: your own data (surveys/interviews) vs existing sources (reports, databases).
● Qualitative vs Quantitative: understand the why vs measure the how many/how much.
● Exploratory → Descriptive → Causal: early discovery → measurement → cause■effect (e.g., pricing impact).
● Attitudinal vs Behavioral: what people say vs what they do (triangulate both).
● Generative vs Evaluative: shape the idea vs test a concrete concept.
4) How To Do Market Research (Step■by■Step)
Step 1 — Frame the decision & objectives
Write a 1■page brief that states the decision to make, hypotheses, and success criteria.
● Decision example: “Should we launch Feature X at PKR 799?”
● Objectives: measure demand, identify target segment, estimate willingness to pay (WTP).
● Hypotheses: H1 users doing workaround Y will show highest intent; H0 no segment difference.
● Success metric: ≥30% ‘very likely to buy’ among target users.
Step 2 — Define target audience & sampling
● Describe segments (demographic, firmographic, psychographic, behavior).
● Jobs-To-Be-Done (JTBD): functional, emotional, social jobs; current alternatives.
● Sampling frame: where these people are (lists, communities, CRM, panels).
● Decide sample size & error tolerance (see cheat sheet).
Step 3 — Choose methods & design instruments
● Exploratory (interviews/focus groups) when the problem is fuzzy.
● Quant (surveys/analytics) when you need measurement & sizing.
● Mixed methods to confirm patterns seen in qual with numbers.
Step 4 — Design great questions (survey/interview)
● Every question must map to an objective (cut questions that don’t).
● Avoid bias: leading (“don’t you love…?”), double■barreled, ambiguous scales.
● Order effects: start broad → specific; sensitive questions last.
● Include one open■ended “why” after key ratings to capture context.
Common question types
Single/multiple choice, Likert (1–5), ranking, MaxDiff (relative importance), open■ended.
Step 5 — Pricing & concept testing (practical)
● Gabor–Granger: show price options one by one; record “would buy?” to find demand curve.
● Van Westendorp (PSM): four questions (too cheap/cheap/expensive/too expensive) to get acceptable price
range.
● Concept tests: show mockups; measure appeal, clarity, uniqueness, and intent to try/buy.
Step 6 — Recruit & fieldwork (quality controls)
● Set quotas (e.g., 60% target segment; balance by city/age as needed).
● Pilot test with 5–10 respondents; fix confusing items.
● Use attention checks, minimum completion time, duplicate detection.
● Log consent, incentives, and source for each respondent.
Step 7 — Clean & analyze data (quantitative)
● Clean: remove straight■liners/speeders; handle missing values.
● Describe: counts, %, means/medians; visualize distributions.
● Compare groups: crosstabs (chi■square), t■tests/ANOVA, simple correlations.
● Segment: rules■based (e.g., heavy users) or clustering when feasible.
● Report uncertainty: confidence intervals, margin of error.
Step 8 — Analyze qualitative data
● Code transcripts: open → axial → selective; create a theme map.
● Build a quote bank to humanize insights.
● Map journeys: stages, pain points, moments of truth.
● Triangulate with survey metrics to strengthen claims.
Step 9 — Synthesize into decisions
● Turn findings into insights: “Because we observed X, it means Y, therefore we should Z.”
● Prioritize with a simple RICE/ICE score; state assumptions & risks.
● Define next experiments (A/B tests, pilots) with owners & KPIs.
Step 10 — Communicate & archive
● Executive one■pager: objective → methods → 3 findings → 3 actions → next experiments.
● Keep a repository: brief, instrument, raw data, codebook, analysis, readout.
● Document limitations and ethical considerations (privacy, consent).
5) Practical Tools (Fast & Low■Cost)
● Surveys: Google Forms, Typeform. Interviews: Zoom/Meet + recording.
● Analysis: Excel/Sheets; pivot tables; basic stats add■ins.
● Desk research: company reports, govt stats, credible industry blogs.
● Market signals: Google Trends; app stores; social listening on brand/keywords.
6) Sizing Your Opportunity (TAM/SAM/SOM)
Top■down: start from a credible market estimate and narrow by segment. Bottom■up: estimate buyers × price
using your target channel coverage. Prefer bottom■up.
Metric Definition Example (Meal■prep App)
TAM Entire addressable market for the need
All urban smartphone users who order prepared food
SAM Segment you can serve with your model
Users in cities you operate, with delivery access
SOM Share you can realistically capture soon
First■year target: 2% of SAM via campus + office corridors
Bottom■up sizing example
Target: 50,000 potential buyers × 10% conversion × PKR 600 avg. order × 1 order/week → ≈ PKR 150M
annualized GMV.
7) Pricing Research — At a Glance
Method What you do Output When to use
Gabor–Granger Ask buy/not■buy across a price ladder
Demand curve; revenue■max
Simple
price
concepts; quick test
Van Westendorp (PSM)
Ask 4 price perception items Acceptable price range Early■stage price feel
Conjoint/MaxDiff (lite)
Trade■off or importance tasks Feature/price utilities When choices are complex
8) Delegation & Workflow (RACI + Cadence)
Role Responsibilities
Research Lead (R/A) Brief, design, quality, synthesis, stakeholder alignment
Sponsor/PM (A) Decision framing, scope, unblockers, approvals
Analyst (R) Cleaning, stats, visuals, repository
Field Ops (R) Recruitment, quotas, incentives, logs
Domain Experts (C) Review instruments, interpret findings
Stakeholders (I) Attend readout, own action items
9) Mini Case Studies (What Good Looks Like)
● Netflix: Secondary trends signaled broadband growth; concept tests confirmed appeal → pivot to streaming →
new TAM unlocked.
● Coca■Cola New Coke: Taste tests won, but emotional brand affinity ignored → launch backlash → holistic
research needed.
● Local Delivery Startup: Surveys + NPS interviews showed reliability mattered more than discounts → invested
in logistics → retention up.
10) Ready■to■Use Templates & Checklists
Research brief (fill this)
● Decision & objectives (SMART) • Background/context
● Target audience & sample size • Methods & timeline
● Success metrics • Assumptions/risks • Stakeholders (RACI)
Interview guide (skeleton)
● Opener & consent • Context (role, goals)
● Current behavior & workarounds • Pain points
● Concept reactions (likes/concerns/suggestions)
● Wrap: 'What would make this a must■have?'
Survey skeleton
● Screener → concept exposure → core need/usage → pricing block → open■end → demographics (last)
Readout one■pager (exec)
● Objective → Methods → 3 Key Findings → 3 Recommendations → 2 Next Experiments (with owners & dates)
Appendix — Sample Size Cheat Sheet (Approx., p=0.5)
Confidence ± Margin n (large population)
95% 10% ≈ 97
95% 5% ≈ 385
95% 3% ≈ 1,067
90% 5% ≈ 271
Pro tip: keep research continuous
Run small, regular studies (monthly pulses) rather than one big study. Momentum beats perfection.
