Research Methods Overview

The Right Tool for the Right Question

In Lesson 1, we established that customer research replaces assumptions with evidence. But which research method should you use? Sending a survey when you need an interview—or running interviews when you need analytics—wastes time and produces the wrong insights.

By the end of this lesson, you’ll be able to:

• Describe the four major categories of customer research
• Choose the right method based on your specific research question
• Design a basic research plan using AI assistance

The Four Pillars of Customer Research

Think of research methods as tools in a toolbox. A hammer is great for nails but terrible for screws. Similarly, each research method excels at answering specific types of questions.

1. Surveys (Quantitative)

What they measure: Attitudes, preferences, satisfaction, and behaviors across large groups.

Best for: “How many customers prefer X?” “What’s our satisfaction score?” “Which features do people want most?”

Strengths: Scalable, statistically significant, easy to analyze.

Weaknesses: Can’t explore “why” deeply; biased by question design; low response rates.

Sample size: 100+ responses for meaningful patterns.

2. Interviews (Qualitative)

What they reveal: Motivations, frustrations, mental models, and unmet needs.

Best for: “Why do customers cancel?” “What’s the decision-making process?” “What problems do we not know about?”

Strengths: Deep insights, uncovers surprises, builds empathy.

Weaknesses: Time-intensive, small sample sizes, interviewer bias risk.

Sample size: 5-15 interviews to reach saturation (when you stop hearing new insights).

3. Analytics (Behavioral)

What they show: What users actually do—page visits, click patterns, conversion funnels, drop-off points.

Best for: “Where do users get stuck?” “Which features are actually used?” “What’s our conversion rate?”

Strengths: Objective (no self-report bias), large sample, continuous data.

Weaknesses: Shows behavior but not motivation; requires proper instrumentation.

Sample size: Depends on traffic, but typically 1,000+ sessions for meaningful patterns.

4. Observation (Contextual)

What it uncovers: How people actually use products in their natural environment—often revealing gaps between what they say and what they do.

Best for: “How do users actually interact with our product?” “What workarounds have they invented?”

Strengths: Uncovers unconscious behaviors, reveals real-world context.

Weaknesses: Time-intensive, observer effect (people behave differently when watched).

Sample size: 5-10 observation sessions.

✅ Quick Check: A customer says they love your product in a survey but their usage data shows they log in once a month. Which research method would reveal the truth—and which would be misleading?

The Research Question First Approach

The most common mistake in customer research is choosing a method before defining the question. Instead, follow this flow:

Step 1: Define your question. What specific decision does this research need to inform?

Step 2: Classify the question.

• “How many/often?” → Quantitative (survey or analytics)
• “Why/how?” → Qualitative (interview or observation)
• “What do they actually do?” → Behavioral (analytics or observation)

Step 3: Choose the method. Match question type to research method.

Step 4: Design the study. Create the instrument (survey questions, interview guide, analytics dashboard).

Decision Matrix

How AI Helps

“I run an online tutoring platform. Signups are strong but only 30% of new users book their first session. I need to understand why. Recommend the best research approach—which methods to use, in what order, and why. Include sample sizes and estimated timelines.”

Building a Research Plan

A research plan keeps your project focused and efficient. Here’s the template:

RESEARCH PLAN
=============
Research Question: [What do we need to learn?]
Business Decision: [What will this research inform?]
Method(s): [Survey / Interview / Analytics / Observation]
Participants: [Who, how many, how recruited]
Timeline: [Start date → Analysis complete]
Deliverable: [Report, presentation, persona, etc.]

How AI Helps

“Help me create a research plan. I need to understand why enterprise customers are churning after 6 months. My constraints: 2-week timeline, budget for 10 interview incentives ($50 each), and access to our analytics dashboard. Design a mixed-method research plan.”

Combining Methods: The Power of Mixed Research

The strongest research combines multiple methods. A common pattern:

1. Start with analytics to identify patterns and generate hypotheses.
2. Run interviews to understand the “why” behind the data.
3. Validate with surveys to confirm findings at scale.

This funnel moves from broad observation to deep understanding to statistical confirmation.

Try It Yourself

Think of a question about your customers (or a hypothetical business) and use AI to plan the research:

“I want to understand [research question]. My business is [description]. I have [timeline] and [budget/resources]. Help me choose the best research method(s) and create a basic research plan with specific steps.”

Key Takeaways

• Four main methods: Surveys (how many), interviews (why), analytics (what they do), observation (how they do it)
• Start with the question, not the method—the question determines which tool works best
• Mixed-method research (combining quantitative + qualitative) produces the strongest insights
• A research plan keeps your project focused, efficient, and actionable
• AI helps design research approaches and create professional research plans in minutes

Up Next

In Lesson 3: Designing Effective Surveys, we’ll go deep on the most widely used research method. You’ll learn to write unbiased questions, structure surveys for high completion rates, and use AI to catch common survey design mistakes.