Product research key terms.
40 terms every PM, UX researcher, and product-minded developer should know — from affinity mapping to voice of customer.
Affinity Mapping
A collaborative synthesis technique that groups raw research observations, quotes, or ideas into thematic clusters to surface patterns. Teams physically or digitally arrange notes until natural groupings emerge, revealing shared pain points, mental models, and opportunities that might not be visible in individual data points.
How Evermuse automates thematic grouping →Ask Anything
A natural-language query interface that lets product teams interrogate their entire repository of customer conversations without manual tagging or search syntax. Ask questions like "What are customers saying about our onboarding?" and receive cited, evidence-backed answers drawn from real meeting recordings and transcripts.
See Ask Anything in Evermuse →Churn
The rate at which customers stop using a product over a given period. Churn is a lagging indicator — by the time it appears in metrics, the underlying causes (unmet needs, friction, competitor switching) are already embedded in customer conversations. Analyzing those conversations before customers leave is how proactive teams reduce churn.
Continuous Discovery
An ongoing practice — popularized by Teresa Torres — of maintaining weekly contact with customers through frequent, lightweight research sessions rather than occasional large research sprints. The goal is to keep a living picture of customer needs constantly updated so product decisions are always grounded in fresh evidence, not six-month-old research.
How Evermuse supports continuous discovery →Customer Effort Score (CES)
A metric that measures how much effort a customer had to exert to accomplish a goal, typically on a scale of 1–7. CES is particularly predictive of customer loyalty: high effort correlates strongly with churn and negative word-of-mouth. It's most commonly deployed after support interactions or onboarding touchpoints.
Customer Journey
The full sequence of touchpoints and experiences a user has with a product or brand — from first awareness through purchase, onboarding, ongoing use, and potential churn. Journey mapping makes invisible friction visible by documenting what customers do, think, and feel at each stage, revealing where the product falls short of expectations.
Customer Satisfaction Score (CSAT)
A simple metric that measures how satisfied customers are with a specific interaction, feature, or experience, typically on a 1–5 or 1–10 scale. CSAT captures point-in-time satisfaction rather than overall loyalty; it's most valuable when combined with open-ended qualitative feedback that explains the score.
Customer Segment
A distinct group of customers who share common characteristics, needs, behaviors, or contexts relevant to product strategy. Segments aren't always demographic — they're often defined by job function, use case, maturity, or problem-type. Effective segmentation helps teams avoid building one-size-fits-all solutions that actually fit no one well.
Discovery
The research phase of product development focused on identifying which customer problems are worth solving before committing to solutions. Discovery involves interviews, observation, and synthesis to understand what customers are trying to accomplish, what gets in their way, and what an ideal solution would look like — all before a line of code is written.
Discovery Sprint
A concentrated, time-boxed period — typically one to two weeks — dedicated to answering a specific product question through targeted customer research. Unlike continuous discovery, sprints are episodic and appropriate for exploring a new problem space, evaluating a strategic bet, or de-risking a major product decision.
Evidence-Based Roadmap
A product roadmap in which every item is explicitly tied to customer evidence — verbatim quotes, recurring themes, or quantified need frequency — rather than internal assumptions or stakeholder preferences. Evidence-based roadmaps shift the conversation from "I think" to "customers told us," making prioritization more defensible and reducing the risk of building for hypothetical users.
See how Evermuse grounds roadmaps in evidence →Feature Request
A customer suggestion for new functionality or a change to existing behavior. Feature requests are raw product signals — they reveal surface wants but rarely the underlying need. The job of product research is to look behind feature requests to understand the problem the customer is actually trying to solve, which may have a better solution than the one they suggested.
Friction
Any element of a product experience that makes a task harder, slower, or more confusing than it needs to be. Friction is a major driver of churn and abandonment, and it's best discovered through usability testing and customer interviews — users will work around friction silently rather than complain about it.
Hypothesis Testing
A structured approach to validating or invalidating assumptions about customer behavior, product value, or market fit. A good hypothesis is falsifiable, time-bound, and tied to a decision — "if X is true, we'll do Y; if false, we'll do Z." Testing replaces gut feel with evidence and reduces the cost of being wrong by catching bad bets early.
Insight
A non-obvious, actionable understanding about customers that changes how a team thinks or what it decides to build. Raw data (quotes, observations, metrics) becomes an insight only through synthesis — when patterns across multiple sources reveal something meaningful that wasn't visible in any single data point.
Insight Repository
A centralized, searchable database of synthesized learnings from customer research. A good repository is living (continuously updated), tagged (organized by theme, customer, product area), and evidenced (every insight links back to the original source). Without a repository, research knowledge is trapped in docs and Slack threads and decays rapidly as teams grow.
How Evermuse works as your always-on insight repository →Jobs-to-be-Done (JTBD)
A framework — developed by Clayton Christensen — for understanding the fundamental goal a customer is trying to accomplish when they use a product. Customers "hire" products to do a job; understanding the job (not just the feature request) helps teams build solutions that actually satisfy the underlying need. The classic example: people don't want a drill, they want a hole in the wall.
MCP (Model Context Protocol)
An open standard that allows AI coding assistants and language models to connect to external tools and data sources. Through MCP, an AI like Claude or Cursor can directly query a product team's customer evidence, roadmap, and research notes — bridging the gap between developer tools and product knowledge.
Explore the Evermuse MCP →Mixed Methods Research
A research approach that combines qualitative methods (interviews, observation) and quantitative methods (surveys, analytics) to get a fuller picture of customer behavior. Qualitative data tells you what is happening and why; quantitative data tells you how often and at what scale. The best product teams use both together.
Net Promoter Score (NPS)
A metric that measures customer loyalty by asking "How likely are you to recommend this product?" on a 0–10 scale. Respondents are grouped into Promoters (9–10), Passives (7–8), and Detractors (0–6). NPS = % Promoters − % Detractors. The score is a useful pulse check, but its real value comes from the open-ended follow-up that explains the rating.
Opportunity Scoring
A prioritization framework developed by Anthony Ulwick that quantifies where to invest product development by measuring how important an outcome is to customers vs. how satisfied they are with existing solutions. High importance + low satisfaction = underserved opportunity. Low importance + high satisfaction = overserved, don't invest more.
Pain Point
A specific problem, frustration, or obstacle that a customer encounters in their current workflow or with an existing product. Pain points are the raw material of product discovery — understanding which pains are most common, most severe, and most addressable is how teams decide where to focus. Verbatim customer quotes are the most credible evidence of real pain.
Product Intelligence
The systematic collection, analysis, and synthesis of signals from customers, market trends, and product usage to inform strategic decisions. Product intelligence closes the loop between what customers say (voice of customer), what they do (analytics), and what they need (discovery) — giving teams a complete, continuously updated picture of reality.
See how Evermuse delivers product intelligence →Product Signal
Any piece of evidence — a quote from a customer interview, a support ticket theme, a usage pattern, a churn reason — that reveals something meaningful about customer needs or product performance. Individual signals are noisy; their value comes from aggregation and synthesis across dozens or hundreds of conversations.
Qualitative Research
Research methods that produce non-numerical, contextual insights: user interviews, observational studies, focus groups, and open-ended survey responses. Qualitative research excels at explaining why customers behave the way they do and surfaces surprising context that quantitative data alone can never reveal. Its limitation is that it's hard to scale — synthesis is time-intensive.
How Evermuse scales qualitative research →Quantitative Research
Research methods that produce numerical data: structured surveys, usage analytics, A/B tests, and funnel analysis. Quantitative methods are strong at showing what is happening at scale but rarely reveal why. A drop in activation rate is a quantitative finding; understanding the friction causing it requires qualitative investigation.
Research Repository
A structured library of past research studies, interview transcripts, synthesis artifacts, and insights that a product organization can reference, search, and build upon over time. A well-maintained repository prevents teams from re-interviewing the same customers, losing institutional knowledge when researchers leave, and making decisions that contradict prior findings.
How Evermuse builds your research repository automatically →Research Sprint
A concentrated, time-boxed research effort — typically one to two weeks — focused on answering a specific product question. Sprints follow a defined process: frame the question, recruit participants, conduct sessions, synthesize findings, share results. They're most effective when the question is sharp and the findings feed directly into an upcoming product decision.
Signal-to-Noise Ratio
In the context of customer feedback, the proportion of meaningful, actionable insights versus irrelevant, redundant, or misleading data. Raw customer feedback — especially at scale — is dominated by noise: one-off edge cases, venting without context, and feature requests that mask deeper needs. Good synthesis dramatically improves the signal-to-noise ratio.
Synthesis
The analytical process of examining multiple research data points — quotes, observations, notes — to identify patterns, themes, and actionable insights. Synthesis transforms raw evidence into understanding. It's the hardest and most valuable part of the research process, and the step most often skipped when teams are under time pressure.
How Evermuse automates synthesis →Thematic Analysis
A qualitative research method for identifying, analyzing, and reporting recurring patterns (themes) within a data set. In product research, thematic analysis applied to interview transcripts surfaces the dominant customer needs, pain points, and mental models — providing the structural backbone of an insight repository.
See how Evermuse runs thematic analysis automatically →Transcript
A verbatim text record of a spoken conversation, user interview, or customer meeting. Transcripts are the primary raw material for qualitative product research — they preserve the exact words customers used, the questions asked, and the context of the conversation, enabling synthesis and evidence-backed insights that summaries alone cannot support.
Usability Testing
A research method that observes real users attempting to complete specific tasks with a product — either a prototype or the live product. Unlike user interviews (which explore needs and context), usability tests directly reveal where the interface confuses, frustrates, or blocks users. Even five participants can uncover the majority of major usability problems.
User Interview
A one-on-one conversation between a researcher and a current or potential customer, designed to understand their context, goals, workflows, behaviors, and frustrations. User interviews are the foundation of qualitative product research. When conducted and synthesized consistently, they are the most reliable source of customer truth in a product organization.
How Evermuse turns every interview into lasting insight →User Journey Map
A visual representation of the steps a user takes to accomplish a goal, annotated with their thoughts, feelings, and pain points at each stage. Journey maps are synthesized from research — not invented — and are most useful when they reveal unexpected breakdowns and emotional low-points that the team hadn't previously recognized.
User Persona
A research-based, fictional representation of a key user segment, capturing their goals, behaviors, context, frustrations, and mental models. Personas are a communication artifact, not a research deliverable — their purpose is to align a team around who they're building for. Personas built from actual interviews are valuable; personas invented in a workshop are often misleading.
User Story
A short description of a desired product behavior written from the perspective of the user, in the format: "As a [user type], I want [action] so that [outcome]." User stories are a collaboration tool between product and engineering, not a research output. The best user stories are informed by real user interviews — the "so that" clause should reflect an actual customer need.
Verbatim
The exact words spoken or written by a customer, preserved without paraphrase or interpretation. Verbatim quotes are the gold standard of customer evidence because they cannot be unconsciously distorted by researcher bias. A roadmap decision backed by ten verbatim quotes from real customers is far harder to dismiss than one backed by a summary or a gut feeling.
Voice of Customer (VoC)
The direct capture of customers' own words, expectations, preferences, and frustrations — as opposed to inferences drawn from their behavior. VoC programs gather data from interviews, support conversations, reviews, surveys, and sales calls, then synthesize it to surface the themes that most influence product strategy. The goal is to hear customers, not just measure them.
How Evermuse captures your voice of customer →Win/Loss Analysis
A systematic review of won and lost sales deals to understand why customers chose your product or a competitor's. Win/loss interviews — conducted with recent prospects regardless of outcome — reveal competitive positioning gaps, messaging failures, and product shortcomings that internal teams rarely see. Sales and CS conversations are an underused source of this signal.
How Evermuse captures product signals from sales conversations →