How we work
Methodology and Privacy
How BeMoreKinky research data is collected, anonymised and analysed, including the collection's privacy architecture, definitions and analytical methods.
BeMoreKinky research starts with things people do in the app: rating an activity, answering a profile question or responding to a partner's proposal. The research data contains anonymous selections and outcomes, which we turn into percentages.
When we write about couples, we compare the two sides of the same activity. "Blindfold them", for example, is matched with "be blindfolded". Private chat messages and journal entries never enter the analysis.
On this page
What is BeMoreKinky? · What app activity goes into the research? · What do acceptance, mutual yes and conflict mean? · Who do the findings reflect? · How does BeMoreKinky keep private writing out of the research? · How do we turn app activity into percentages?
What is BeMoreKinky?
BeMoreKinky is an app for couples to talk about BDSM and kink. Partners can rate activities separately, compare the things they both want, plan play, set tasks and take profiles or quizzes.
Across the research collection, we draw on activity from 25k+ couples. Each article uses the part of the app that answers its question. A piece about partner agreement, for example, needs answers from both partners. A piece about quizzes needs completed quiz answers.
What app activity goes into the research?
The research collection draws on five kinds of app activity:
- Activity preferences: yes, no or maybe on a particular activity.
- Partner comparisons: answers from both people on the giving and receiving sides of that activity.
- Proposal outcomes: whether a suggested play plan was accepted, declined or left unanswered.
- Profiles and quizzes: the choices people made while completing them.
- Scene and task features: anonymous events created as people used those parts of the app.
Most of the current collection covers activity recorded between October 2024 and February 2026.
Before analysing a question, we choose the source that can answer it. A preference says what somebody selected. A partner comparison tells us whether two linked answers match. A proposal outcome tells us what happened to a proposal in the app. They are different kinds of evidence, so we keep them separate.
What do acceptance, mutual yes and conflict mean?
| Term | What it means in our research |
|---|---|
| Acceptance rate | Of everyone who answered an activity, the percentage who chose yes. |
| Mutual yes | Both partners chose yes on the paired prompts: one to giving the activity and the other to receiving it. |
| Conflict | One partner chose yes and the other chose no on those paired prompts. |
| Concordance | The percentage of paired answers that match: yes with yes, no with no or maybe with maybe. |
| Experience level | The level somebody chose for themselves in the app, from Level 1 to Level 4. |
The direction of a prompt matters. "Blindfold them" and "be blindfolded" form a pair, but they ask different people different questions. Individual preference results keep that direction visible. Couple results match the two directions before we calculate a percentage.
If somebody did not answer a prompt, we do not count the missing answer as a no.
Who do the findings reflect?
The findings describe people who chose a kink app and used the feature being studied. An activity result refers to people who answered that activity; a profile result refers to people who completed that profile; a couple result refers to partners who both answered the relevant pair. We use the same specific language in each article rather than treating one app result as a claim about everyone who practises kink.
How does BeMoreKinky keep private writing out of the research?
The research data is organised around anonymous identifiers and event types, not names or email addresses. Anonymous partner identifiers let us connect the two sides of an activity without putting either person's identity into the analysis.
We analyse selections and feature outcomes, not intimate conversations. Private chat messages and journal text sit outside the research path. To find out whether partners agreed on blindfolding, we need their two anonymous answers. We do not need the conversation that led to them.
The published articles show percentages. The underlying respondent and event totals are not made public.
How do we turn app activity into percentages?
How do we calculate acceptance rates?
For an activity, we calculate the percentage of completed answers that were yes. When the question concerns a particular group or direction, such as people answering the receiving prompt, we calculate the percentage within that named group.
How do we match partners' answers?
We link each giving prompt to its receiving counterpart. A couple enters the comparison only when both partners answered the relevant side. Their paired answers can then be counted as a mutual yes, a conflict or part of the concordance rate.
Two activities can each be popular without being a match between partners. Pairing the prompts is what lets us answer the couple-level question.
How do we compare experience groups?
People choose an experience level from Level 1 to Level 4. We calculate the relevant percentage within each level, then compare those groups.
How do we measure proposals, tasks, profiles, quizzes and scenes?
For proposals, tasks, profiles, quizzes and scene features, we define the event that answers the question before calculating a percentage. A proposal article might compare proposal outcomes. A profile article might compare how often people agreed with each statement. Unlike events stay on separate scales.
How do we check results before publication?
Before publishing a result, we check its percentage against the source and make sure the sentence names the right group and prompt direction. We also check that the prose, figure and accessible description all show the same rounded value.
Want to ask us about the research method?
For questions about the method or academic use of the research, contact the BeMoreKinky team.