Conversion Research Methods: The Complete Toolkit

User Research for CRO: Why Data Alone Isn’t Enough

Analytics tells you what users do. User research tells you why. The best CRO programs combine both to generate hypotheses that actually win.

Quantitative Research

Analytics Analysis

Funnel drop-off identification
Segment-based conversion analysis
Page-by-page performance review
Source/medium attribution
Time-based pattern analysis

Heatmaps

Click maps: Where users click (and don’t)
Scroll maps: How far down pages users scroll
Move maps: Where attention focuses
Confetti reports: Individual click visualization

Form Analytics

Field-by-field abandonment
Time spent per field
Error rate per field
Re-correction patterns

Qualitative Research

User Interviews

When: 5-8 interviews per segment provides 80% of insights

Format: 30-45 minute video calls, recorded

Topics:

Their journey to finding you
What almost stopped them from buying
What convinced them
What they expected vs experienced
Words they use to describe your product

Customer Surveys

On-Site Surveys:

Exit-intent: “What stopped you from purchasing today?”
Post-purchase: “What almost prevented you from buying?”
Page-specific: “Was this page helpful?”

Email Surveys:

NPS: “How likely to recommend?”
CSAT: “How satisfied are you?”
Open-ended: “What’s the #1 thing we could improve?”

Session Recordings

Review 50-100 sessions per month focused on:

Drop-off points
High-intent users (added to cart, didn’t buy)
Confused behaviors (rage clicks, repeated scrolling)
Successful conversions (what worked)

Usability Testing

5 users uncover 80% of usability issues
Task-based testing (not free exploration)
Think-aloud protocol
Compare with competitor experiences

Voice of Customer Analysis

Sources

Customer service tickets and chat logs
Product reviews (yours and competitors’)
Social media mentions and comments
Sales call recordings
Survey open-ended responses

What to Extract

Their language — use these words in copy
Their objections — address these on key pages
Their goals — lead with these benefits
Their alternatives — understand competitive context
Their journey — map all touchpoints they mention

Research-to-Test Pipeline

Conduct research (mix of quant + qual)
Identify patterns across data sources
Generate hypotheses based on findings
Score and prioritize with ICE/AXR
Test the highest-priority hypotheses
Document learnings and inform next round of research

Building a Research Operations Practice

Research ops makes user research repeatable, organized, and accessible across the company. Without it, research becomes one-off projects that don’t compound.

Core Components

Participant database: Recruited customers segmented by attributes (industry, plan, use case, recency)
Research repository: Searchable storage of all research findings with tags and themes
Standardized templates: Interview guides, survey scripts, usability test protocols
Tooling stack: Recruiting (User Interviews), conducting (Zoom/Loom), analyzing (Reduct, Dovetail)
Cadence calendar: Quarterly research planning aligned to product roadmap

Research Repository Structure

Tags: Page tested, persona, journey stage, finding type
Themes: Recurring patterns synthesized across studies
Highlights: Direct user quotes with timestamps and context
Insights: Synthesized conclusions with actionable implications
Artifacts: Recordings, transcripts, screenshots, prototypes

Recruiting Quality Research Participants

The single biggest determinant of research quality is who you talk to. Bad recruits produce useless insights.

Sources for Customer Research

Customer database: Recent buyers, high-value accounts, segments matching study criteria
Email list: Opted-in subscribers willing to provide feedback
In-app intercept: Trigger survey or research invitation based on behavior
NPS/CSAT respondents: People who’ve already shown willingness to share feedback
Customer advocacy program: Power users who want to influence product

Sources for Prospect Research

User Interviews / UserTesting: Pre-screened panels with demographic targeting
Respondent.io: Higher-quality professional panels for B2B research
LinkedIn outreach: For specific roles/industries
Reddit/community recruiting: For niche audiences
Trade event recruiting: Conference attendees in your space

Incentive Best Practices

B2C: $50-100 for 30-60 minute interviews
B2B professionals: $150-300 depending on seniority
Executive interviews: $500+ or charity donation
Existing customers: Premium product access, swag, or modest gift cards
Pay promptly via Tremendous, Rybbon, or PayPal

Research Methods Selection Matrix

Different questions require different methods. Don’t default to surveys for everything.

Question Type	Best Method	When to Use
What do users want?	1:1 interviews	Discovery, persona research
Why did this happen?	Session recordings + interviews	Diagnosing drops, friction
How do users do X?	Usability testing	Pre-launch validation
How many feel Y?	Survey at scale	Quantifying themes
What are users doing?	Analytics + heatmaps	Behavioral patterns
Which version is better?	A/B test or preference test	Decision validation
What words do users use?	Customer interviews + reviews	Copy and messaging research

Synthesizing Research Into Insights

Raw research is useless. The value comes from synthesis — patterns that lead to action.

The Synthesis Process

Transcribe everything: Use Reduct, Otter, or Rev for searchable transcripts
Highlight key moments: Tag direct quotes that reveal pain, joy, confusion, or workarounds
Cluster by theme: Group similar quotes across multiple participants
Identify patterns: Note when 3+ people independently mention something
Frame as insights: Convert observations into actionable insights
Tie to opportunities: Connect each insight to a product or marketing opportunity

Insight Templates

Pain insight: “Users struggle with X because Y, leading to Z”
Mental model insight: “Users expect X to mean Y, but it actually means Z”
Workaround insight: “Users currently solve X by doing Y, which suggests they need Z”
Language insight: “Users describe X as Y in their own words”
Comparison insight: “Users compare us to X, evaluating us on Y criteria”

Common Research Mistakes

1. Leading Questions

Bad: “How frustrating is our checkout?”

Good: “Tell me about the last time you bought online — walk me through it.”

2. Confirmation Bias

Going in with hypotheses you want validated rather than questions you want answered. Mitigate by writing down what you expect to hear, then noticing when reality differs.

3. Talking Too Much

The 80/20 rule: participants should talk 80%, you should talk 20%. Embrace silence — people fill silences with the most valuable information.

4. Recruiting From One Segment

A dozen interviews with power users won’t tell you anything about new users’ struggles. Diversify recruiting across the user lifecycle.

5. Synthesis-Free Research

Conducting research without dedicating time to synthesize is wasted effort. Allocate 2-3x the interview time to analysis.

Frequently Asked Questions

How many interviews until I see patterns?

For most B2C research, 5-7 interviews per segment reveal 70-80% of major patterns. B2B with greater variance often requires 8-12 per segment. Stop when you start hearing the same insights repeatedly.

Should I record research sessions?

Yes — always (with permission). Transcripts and recordings allow you to revisit sessions, share quotes with team members who weren’t present, and build searchable repositories. Modern tools make this trivial.

What’s the difference between UX research and CRO research?

UX research focuses on usability and user experience broadly. CRO research focuses specifically on conversion barriers and motivators. They overlap heavily — the difference is more about which questions you prioritize.

How do I get stakeholders to actually use research findings?

Involve them in the research itself. Stakeholders who watch interviews live become research advocates. Stakeholders who only read reports treat findings as someone else’s opinion.