Game user research is the practice of studying how real players experience a game, so the team building it can make better decisions. It covers everything from watching a single new player try a tutorial to running a survey across thousands of people in a live game. The discipline is also called games user research, or GUR, and it is often a small team, or one researcher, working between design, product, and UA.

This page defines the discipline, separates the three kinds of evidence it works with, and then organizes it around the five decisions it most often serves: onboarding, difficulty, progression, monetization, and live events. The other guides in this series go deeper on each method and question.

What game user research is

A game user researcher answers questions about players. Where do new players get lost? Is level 12 too hard, or is it hard in the way players enjoy? Does the shop make sense to someone who has never seen it? Will players come back for a weekend event?

The work borrows its methods from usability research, psychology, and market research, and adapts them to something those fields were not built for. A game is meant to be difficult in places. A banking app should never make people struggle. A puzzle game has to. Game user research has to tell the two kinds of struggle apart, and that shapes almost every method the discipline uses.

Game user research vs. market research

Market research for games asks who the audience is, how big it is, and what they say they want. It tends to work before a game exists or outside it: genre surveys, concept polls, competitor analysis, and store data.

Game user research works with the game itself, or a prototype of it, in front of players. The two overlap at concept testing, where a team shows players an idea before building it. After that point, market research tells you who to recruit, and user research tells you what happened when they played.

What a playtest is

A playtest is a session where a player plays the game while someone studies what happens. Most teams use the word that way, but in practice a playtest can be one player in a lab with a researcher beside them, 30 players at home on their own phones, or a closed beta with thousands of people.

Playtesting is one method inside game user research, and the most common one. The discipline also includes surveys, interviews, diary studies, and the analysis of telemetry, which is the event data a game logs as people play. Playtesting methods: which one answers which question covers each of these in detail.

Three kinds of evidence

Every game user research method produces one of three kinds of evidence, or some mix of them.

What players say

Surveys, interviews, ratings, and think-aloud comments all record what players say about their experience. This is the easiest evidence to collect and the easiest to act on, because it arrives already in words.

It is also the evidence most shaped by memory and by the situation. Players forget the middle of a session and remember the most intense moment and the ending, a pattern psychologists call the Peak-End Rule. They soften criticism when the person asking built the game. They explain decisions after the fact in ways that sound reasonable and may not match what drove them. Across our own studies, 50% of in-game emotional responses never appear in what players say in surveys or interviews. What players say is still worth collecting. It tells you what they noticed, what they value, and how they will describe your game to a friend.

What players do

Telemetry, funnel data, heatmaps, and session recordings record what players do. This evidence is precise about location. Analytics can tell you how many new players never finish level 3, and exactly where each of them stopped.

It cannot tell you why. A player who stopped because they were confused and a player who stopped because they were bored leave the same trace in the event log. Analytics or playtesting for Day-1 churn? walks through an example where two players produced the same event trace for different reasons.

What a researcher observes during play

The third kind of evidence comes from watching play as it happens: a researcher noticing where a player hesitates, where they lean in, where they stop reading the dialogue and start tapping through it. Observation sits between the other two. It sees behavior in context, which analytics cannot, and it does not depend on the player's memory, which self-report does.

Human observation has limits too. A researcher can only watch one thing at a time, and some reactions happen too fast to see.

At Emhance, we record participants through a standard webcam while they play at home, and analyze their facial expressions and eye movement second by second. We count a moment as engaged only when emotional intensity and concentration are both high, because either signal on its own can mean something else. You can try the blink signal behind concentration on your own webcam in our 30-second demo game.

Most good studies combine all three kinds of evidence. A typical session might pair a short interview before play, free play while concentration and emotional intensity are measured, and a debrief afterward that asks about the moments the data flagged.

What teams use game user research for

Studios usually commission research because a decision is coming and someone wants evidence before making it. Most of it serves one of five decisions.

Onboarding

The question: do new players understand the game well enough, and enjoy it soon enough, to come back tomorrow?

Onboarding covers the first-time user experience (FTUE), usually the first five to 15 minutes. Industry data puts mobile game churn at 97% of users by Day 30. FTUE: what the first-time user experience means in games goes through this window in detail.

Research here tends to use small playtests with players who have never seen the game, because nobody on the team can see the tutorial with fresh eyes anymore. Peaksel’s analytics showed a sharp drop before level 3 of The Vault. In a six-player study, we found the tutorial had taught players to work one vault face at a time, while the live game expected them to explore several faces at once. Engagement dropped within the first 45 seconds of live play. Peaksel redesigned levels 1 to 3, and saw 9.6% more users completing level 3 and a 14.5% total LTV uplift. The Peaksel case study covers the study and the redesign in full.

Difficulty

The question: is this hard in a way players enjoy, or hard in a way that makes them leave?

Difficulty research looks for the difference between challenge and confusion. Both slow players down. Both can show up as a failed level in analytics. Challenge keeps players engaged and retrying. Confusion means they do not know what the game wants from them.

In our Royal Match study, levels were most engaging when players were 60% to 80% of the way through, with the highest emotional intensity in the final third. A single failure usually raised emotional intensity, and players retried with more focus. Repeated failure was what eventually brought engagement down. One participant said: "When I feel like I'm so close but I have so few moves left, that's when I'm really engaged."

The full study is in our Royal Match deconstruction.

Studies on difficulty usually need participants who match the target player closely, since an experienced genre player and a newcomer can react to the same level in opposite ways.

Progression

The question: do players want what the game is asking them to work toward?

Progression research covers rewards, upgrades, unlocks, and metagames. It often uses longer studies, because a progression system only shows its shape after several sessions. Diary studies and repeat playtests are common here.

In our Merge Mansion study, all five participants responded more strongly to searching for and building an item than to the moment they handed it over and watched the estate change. The reward screen was one of the quietest parts of the session. The handover still mattered, because it turned a few minutes of board activity into a permanent change, and each finished task revealed the next unfinished one. Deconstructing Merge Mansion's retention engine covers the full study.

Monetization

The question: when and how should the game ask players to spend?

Monetization research tests offer timing, price points, store layout, and how clearly players understand what they are buying. Much of this work is done with A/B tests in a live game, which show which version earned more. Research adds the part an A/B test leaves out: what the player was experiencing when the offer appeared.

Our Monopoly GO! study found zero monetization offers on Day 1. The game let players build their first Landmark before any purchase prompt appeared, so the first offers arrived after players already had something they wanted to finish. Participants also named a price under $2 as their comfort zone, below the game's cheapest offer at $2.99. Our Monopoly GO! deconstruction walks through the first five days of offers.

Live events

The question: will this event bring players back, and will they understand it when they arrive?

Live events add new rules, currencies, and screens to a game players already know. Research here often tests event tutorials, the clarity of rewards, and how many events can run at once before players feel overwhelmed. Experienced players are usually the right participants, since they are the ones the event is meant to retain.

In our Royal Match study, experienced players found the number of simultaneous events overwhelming. One counted them out loud: "one two three four five six seven different things going on right now." Players also skipped event tutorials and did not notice when they had earned event items.

How the methods fit together

No single method covers all five decisions. A rough map:

The guides in this series go deeper on each part of that map:

• Playtest questions that produce answers you can act on

• Playtesting methods: which one answers which question

• Remote playtesting without a lab

• Moderated or unmoderated: choosing for games

• FTUE: what the first-time user experience means in games

• D1, D7 and D30: what each retention window tells you

• How many playtesters do you need?

• Analytics or playtesting for Day-1 churn?

Emhance runs remote playtests for mobile game studios, with participants from our gamer panels or your own players, and second-by-second engagement data for every session. See how Emhance playtesting works.