- DOI
- 10.5281/zenodo.20229746
Non-Monogamy Is Not One Appetite
Want the practical version? This is the research version: methods, limits, tables, and the argument underneath the data. If you want fewer statistics and more couple-facing guidance, read Types of Non-Monogamy: Which Is Right for You?: it turns the findings into plain-language questions about which version you want, which version your partner wants, and which scripts should stay off the table.
Abstract
BeMoreKinky is a couples-focused application for learning about kink, BDSM, consent, and sexual communication. This paper analyzes anonymized BeMoreKinky application data to ask whether users treat non-monogamy-adjacent kink as a single appetite or as a family of distinct preference regimes.
Across the focal activity-preference analysis, responses did not form one coherent dimension. Private, partner-contained exhibitionism had broad acceptance: 54.0% yes and 72.1% yes-or-maybe. Group play had lower firm endorsement, at 37.0% yes, but the highest major-subcategory maybe rate, at 23.2%, suggesting a negotiation-heavy curiosity zone. Cuckolding was more polarized, with 30.6% yes, 15.0% maybe, and 54.3% no. Cuckolding phrases were the clearest rejection-heavy regime, with 24.5% yes and 66.6% no. A small Voyeurism or Exhibitionism category showed high apparent acceptance, but the analyzed report treats it as sparse and unsuitable for headline claims.
The strongest item-level pattern is not "more non-monogamy" versus "less non-monogamy." The most accepted activities are private, dyadic, and logistically simple: sexy photos, masturbation while watched, private videos, and remote or contained forms of display. The most rejected activities are harsh comparison and humiliation phrases. The most divisive items are partner-sharing and in-room cuckolding scenarios where yes and no rates are nearly balanced. Head-to-head comparison data further complicate the hierarchy: exhibitionism is broadly acceptable but wins less often in forced choices than cuckolding phrases and group play. This supports a model in which broad acceptability, uncertain curiosity, niche salience, and strong rejection are separate dimensions.
These findings argue against treating "interest in non-monogamy" as a single latent trait. They also suggest that maybe responses and divisive items should be read as material for structured negotiation rather than as weak desire or weak refusal.
Evidence Base and Scope
The focal activity analysis covers the five dedicated non-monogamy-related subcategories listed in the analyzed report: Cuckolding, Cuckolding phrases, Exhibitionism, Group Play, and Voyeurism or Exhibitionism. The analyzed preference table contains 324,651 yes/maybe/no response rows across 952 dedicated non-monogamy activity-role rows, spanning 2024-10-29 to 2026-05-15. Subcategory user counts are reported separately below because users can appear in more than one subcategory.
Several constraints matter. These are BeMoreKinky users who engaged with app preference surfaces, not a population sample. Demographic fields have large unknown buckets. Preference ratings are optional and conditional on raters, so they are not equivalent to yes/no/maybe shares. The paper therefore makes claims about structure within this analyzed app dataset, not prevalence in the general public.
Literature Review
Empirical work on consensual non-monogamy (CNM) increasingly shows that "non-monogamy" is not a stable singular category. Perceiver studies find that polyamory, open relationships, and swinging are evaluated differently: polyamory is often rated more favorably than more explicitly sexual forms such as swinging or open relationships, and stigma changes once researchers separate CNM subtypes rather than treating CNM as one bucket (Matsick et al., 2014; Balzarini et al., 2018). Measurement work makes a related point at the level of desire, behavior, and familiarity. In a U.S. Census-based quota sample of single adults, desire for polyamory, lifetime engagement, familiarity with polyamory, and respect for polyamorous people were separate outcomes with different base rates (Moors et al., 2021); attachment avoidance predicted attitudes and willingness but dissociated from current engagement (Moors et al., 2015). The sexual-fantasy literature also treats erotic interest as multidimensional. Fantasy taxonomies find many interests are common without being typical, and paraphilic interest often exceeds lived behavior (Joyal et al., 2015; Joyal & Carpentier, 2017). Latent-class research is especially relevant: modeling arousal and discomfort separately recovers qualitatively distinct classes, including people who are simultaneously aroused and distressed by fantasy content (Canivet et al., 2022). These findings support the present paper's decision to treat yes, maybe, no, and forced-choice salience as separate signals rather than different intensities of one appetite.
A second literature stream explains why the same non-monogamy surface can carry different emotional meanings. Experimental and review evidence shows that consent framing changes appraisals of the same extradyadic act, while jealousy, compersion, and ambivalence are not simple opposites (Mogilski et al., 2019; Balzarini et al., 2021; Rodrigues, 2024). Scale development further decomposes compersion into affective, anticipatory, and erotic factors, showing that even positive responses to a partner's other relationship are not unidimensional (Flicker et al., 2021). This matters for cuckolding and hotwife-adjacent material: among gay men, fantasies of a partner with someone else overlap with voyeurism and group sex, vary by content, and produce positive outcomes only conditionally on personality, attachment, and context; CNM fantasies among monogamously partnered people also show that fantasy salience does not always mean readiness for enactment (Lehmiller et al., 2018; Lehmiller, 2020). Consent scholarship provides the bridge from fantasy to practice. BDSM consent models emphasize explicit negotiation, ongoing agreement, safewords, aftercare, and context-sensitive norms rather than a single yes/no moment (Pitagora, 2013; Williams et al., 2014; Beres & MacDonald, 2015; Tarleton et al., 2025). In this literature, uncertainty is not a weak form of consent; it is an instruction to specify activity, people, setting, language, risk, and aftermath. That framing maps directly onto the app pattern in which group play and public/digital exposure produce elevated maybe rates, while harsh comparison phrases produce strong no responses despite their salience for a subset.
Research Question
The central question is whether BeMoreKinky users treat non-monogamy-adjacent kink as one appetite or as multiple appetites with different acceptance curves.
The practical version of the same question is:
When a person says they are "into non-monogamy," "maybe into non-monogamy," or "not into non-monogamy," what exactly might they mean?
The analyzed data suggest that this question cannot be answered at the category level alone. A person might be strongly open to sending a sexy photo to a partner, unsure about playing with another couple, curious about supervised partner sharing, and repelled by humiliation phrases. These are different erotic and relational problems, not small variations on one theme.
Methods
This paper is a descriptive synthesis. It uses the summary statistics and tables already produced in the activity-preference analysis.
The key measures are:
Yes %: the share of activity responses marked yes.Maybe %: the share marked maybe.No %: the share marked no.Yes-or-maybe %: the share of responses that are not no.Average rating: an optional rating field, interpreted cautiously because ratings exist only for a subset of rows.- Head-to-head win share: directional evidence from forced-choice comparisons.
- User-level demographic contrasts: mean user acceptance compared across experience level, role preference, gender presentation, partnership status, partner gender presentation, couple configuration, and premium status.
Response-level subcategory rates describe activity surfaces. User-level demographic contrasts prevent frequent raters from dominating subgroup comparisons. Unknown demographic buckets are retained because they are large in several fields.
No causal inference is attempted. The paper interprets patterns as app-engagement structures: how users in this analyzed export sorted, accepted, hesitated over, or rejected non-monogamy-adjacent activities.
Results
1. The Subcategory Hierarchy Is Not a Single Continuum
At the broadest level, the five focal subcategories show sharply different yes, maybe, and no profiles.
| Subcategory | Responses | Users | Activities | Yes % | Maybe % | No % | Yes-or-maybe % | Avg rating |
|---|---|---|---|---|---|---|---|---|
| Voyeurism or Exhibitionism | 615 | 24 | 96 | 77.2 | 13.0 | 9.8 | 90.2 | 18.0 |
| Exhibitionism | 104,601 | 4,004 | 310 | 54.0 | 18.1 | 27.9 | 72.1 | 12.3 |
| Group Play | 41,704 | 5,070 | 54 | 37.0 | 23.2 | 39.7 | 60.3 | 14.8 |
| Cuckolding | 117,080 | 2,520 | 192 | 30.6 | 15.0 | 54.3 | 45.7 | 19.0 |
| Cuckolding phrases | 60,651 | 2,372 | 300 | 24.5 | 8.9 | 66.6 | 33.4 | 12.3 |
The apparent hierarchy should not be mistaken for a ranking of "more intense" and "less intense" non-monogamy. Exhibitionism is the stable broad-acceptance category, but its item-level base comes especially from private and partner-contained acts. Group Play is curiosity-heavy: its yes rate is much lower than Exhibitionism's, but its maybe rate is highest among the major stable categories. Cuckolding is more polarized, with more no responses than yes-or-maybe responses, while Cuckolding phrases are the strongest rejection-heavy category despite later forced-choice salience. Voyeurism or Exhibitionism shows the highest apparent acceptance but has only 24 users and 615 responses, so it should be treated as sparse.
2. Role Framing Matters Most Where Language and Humiliation Are Central
The analyzed report separates some activity-role framing into dom and sub roles. The most stable role asymmetry appears in Cuckolding phrases.
| Subcategory | Role | Responses | Users | Yes % | Maybe % | No % | Yes-or-maybe % |
|---|---|---|---|---|---|---|---|
| Cuckolding phrases | dom | 26,454 | 1,184 | 19.2 | 9.3 | 71.4 | 28.6 |
| Cuckolding phrases | sub | 45,596 | 1,905 | 28.6 | 8.6 | 62.8 | 37.2 |
| Cuckolding | dom | 42,524 | 2,225 | 29.6 | 14.9 | 55.5 | 44.5 |
| Cuckolding | sub | 63,925 | 1,605 | 31.5 | 15.1 | 53.4 | 46.6 |
| Exhibitionism | dom | 84,320 | 1,647 | 55.8 | 16.7 | 27.4 | 72.6 |
| Exhibitionism | sub | 54,009 | 2,352 | 52.3 | 19.3 | 28.4 | 71.6 |
| Group Play | dom | 19,820 | 2,760 | 38.4 | 22.7 | 38.9 | 61.1 |
| Group Play | sub | 27,355 | 1,830 | 35.7 | 23.8 | 40.5 | 59.5 |
The role difference is small for Exhibitionism and Group Play. In Cuckolding phrases, however, sub-framed phrases have a 37.2% yes-or-maybe rate compared with 28.6% for dom-framed phrases.
Phrases are not just activities; they are scripted interpretations of the activity. Cuckolding can involve desire, jealousy, pride, surrender, compersion, comparison, erotic shame, or partner display. Phrase content can pin that ambiguity to one meaning, especially when harsh wording turns "my partner with someone else" into "what this means about me."
3. "Maybe" Is a Negotiation State, Not a Measurement Error
If yes and no are treated as the only meaningful answers, Group Play looks merely lukewarm. Its 23.2% maybe rate is analytically distinctive.
The maybe-heavy items identified in the analyzed report cluster around logistical, public, identity-bound, or partner-expanding activities:
| Item | Subcategory | Role | Responses | Yes % | Maybe % | Yes-or-maybe % |
|---|---|---|---|---|---|---|
| Play with another couple | Group Play | sub | 3,040 | 38.3 | 28.1 | 66.4 |
| Perform a private cam show | Exhibitionism | sub | 2,328 | 46.9 | 26.8 | 73.7 |
| Group play session with men and women | Group Play | sub | 1,821 | 35.1 | 26.5 | 61.6 |
| Spend a day at nude beach day | Exhibitionism | sub | 2,910 | 47.3 | 26.3 | 73.5 |
| Post on OnlyFans | Exhibitionism | sub | 1,740 | 21.6 | 26.2 | 47.8 |
| Swing with multiple couples | Group Play | sub | 2,428 | 24.1 | 26.0 | 50.1 |
These are activities where missing information matters. "Play with another couple" depends on the couple, venue, rules, safer-sex agreement, emotional expectations, comparison risks, and exit plan. "Post on OnlyFans" depends on privacy, identifiability, audience, permanence, platform risk, and relationship agreements. In this dataset, maybe is not the absence of preference; it is often the presence of negotiation.
4. The Most Accepted Items Are Private, Dyadic, and Low-Logistics
The highest yes-rate items are concentrated in Exhibitionism, but they are not primarily public or anonymous forms of exposure. They are partner-contained and private.
| Item | Subcategory | Role | Responses | Yes % | Yes-or-maybe % | Avg rating |
|---|---|---|---|---|---|---|
| Receive a sexy photo from your partner | Exhibitionism | dom | 1,632 | 91.5 | 96.5 | 24.0 |
| Tell your partner to masturbate while you watch | Exhibitionism | dom | 2,176 | 91.2 | 96.7 | 14.1 |
| Receive a sexy photo from your partner | Exhibitionism | sub | 2,905 | 87.4 | 95.2 | 18.4 |
| Share a sexy photo with your partner | Exhibitionism | sub | 1,746 | 87.3 | 94.7 | 22.5 |
| Watch a private video from your partner | Exhibitionism | dom | 2,720 | 86.8 | 94.1 | 18.8 |
| Direct wearing a remote-controlled vibrator in public | Exhibitionism | dom | 1,632 | 85.7 | 94.3 | 23.5 |
The broad base of exhibitionism in BeMoreKinky appears to be less about anonymous public exposure than controlled visibility: being seen by a trusted partner, sending or receiving erotic media, watching a partner masturbate, or adding a discreet public/private edge while the couple remains the central audience. In ordinary clinical or cultural language, exhibitionism may evoke public display or an audience outside the relationship; here, the strongest acceptance is better described as consensual partner-directed visibility.
5. The Most Rejected Items Are Harsh Comparison Phrases
The lowest yes-rate items are dominated by harsh Cuckolding phrase content. The analyzed report highlights examples such as:
- "Say: He was more of a man than you" at 8.5% yes and 85.0% no.
- "Say: He didn't hesitate like you do" at 9.2% yes and 84.0% no.
Several dom-framed phrase items sit near 8-13% yes and about 79-85% no. These are not merely unpopular acts. They are interpretive scripts that load the erotic scenario with comparison, inadequacy, hesitation, masculinity, or humiliation.
The distinction matters for both research and practice. A person may be open to partner-sharing fantasy while rejecting a specific humiliation frame. Another may be aroused by humiliation language but not want actual partner sex with another person. Another may enjoy hotwife pride and partner desirability while rejecting degradation entirely. "Cuckolding" is therefore too broad a label unless it is broken into emotional tone, proximity, sexual boundary, partner selection, language, and aftercare.
6. Divisive Items Are More Informative Than Unpopular Items
The most divisive activities are not the same as the least popular activities. The analyzed report defines divisive items as those that balance yes and no rather than merely attracting many no responses. Strong examples include:
| Item | Pattern |
|---|---|
| Facilitate partner sharing (temporary, supervised) | 39.2% yes and 39.2% no |
| Watch from a chair in the same room | 42.0% yes and 41.7% no |
| Maintain eye contact with partner during encounter | 42.9% yes and 41.5% no |
These items expose fault lines inside non-monogamy-adjacent kink. They are close enough to the fantasy center to attract substantial yes responses, but concrete enough to activate refusal. They involve proximity, witnessing, partner attention, and the meaning of presence: eye contact can become reassurance, dominance, humiliation, compersion, or pressure depending on the couple. The item-level split suggests that these micro-details are not decorative; they are the kink.
7. Swipe Acceptance and Forced-Choice Salience Diverge
Head-to-head comparisons provide a different signal from yes/maybe/no swipes. The analyzed report treats this evidence as directional because DS-4a contains 8,485 comparison events from only 196 users, mostly brute-force mode. Still, the contrast is conceptually important.
| Subcategory | Offered | Wins | Win per offer % |
|---|---|---|---|
| Cuckolding phrases | 3,012 | 2,015 | 66.9 |
| Group Play | 1,475 | 465 | 31.5 |
| Cuckolding | 2,940 | 1,868 | 63.5 |
| Exhibitionism | 6,830 | 1,650 | 24.2 |
This partly reverses the swipe hierarchy. Exhibitionism has broad yes-or-maybe acceptance, but it wins less often per offer in forced comparisons. Cuckolding phrases are broadly rejected, but perform better per offer among comparison events. Group Play also performs well despite lower yes rates than Exhibitionism.
The simplest interpretation is that users have at least two separable preference dimensions:
- Acceptability: "Would I do this, consider this, or allow this into the conversation?"
- Salience: "When choosing between two possibilities, does this one pull me more strongly?"
Private exhibitionism may be widely acceptable because it is familiar, dyadic, and low-logistics. Cuckolding phrases may be rejected by many users but highly charged for a smaller group. A global yes-rate ranking will over-recommend broadly acceptable items and under-detect niche items that matter intensely to a subset; a salience-only ranking could overstate polarizing content. Good interpretation needs both.
8. Experience Level Is the Clearest User-Level Contrast
The analyzed report finds that most demographic effects are small. The strongest observed user-level contrast is experience level:
| Axis | Contrast | n | Mean yes % | Effect size |
|---|---|---|---|---|
| Experience level | Level 4.0 vs level 2.0 | 1,116 vs 780 | 47.5 vs 30.9 | d = 2.52 |
| Partnership status | Multiple vs single | 250 vs 108 | 54.4 vs 38.2 | d = 1.71 |
| Partner gender presentation | Both vs masculine | 344 vs 1,375 | 46.9 vs 38.2 | d = 1.60 |
| Gender presentation | Both vs feminine | 570 vs 951 | 45.9 vs 39.4 | d = 0.72 |
| Role preference | Switch vs undecided | 855 vs 596 | 45.1 vs 39.3 | d = 1.05 |
| Premium status | Premium vs free | 6,220 vs 219 | 42.6 vs 39.4 | d = 0.44 |
Eta-squared values reinforce the same point. Experience level is the only axis above roughly 0.12, at eta^2 = 0.155; the other axes are smaller in this analyzed dataset.
This should not be overread. Experience level may reflect comfort, self-selection, app tenure, premium engagement, prior kink exposure, or survivorship among people who continue rating more advanced content. Still, the contrast suggests that non-monogamy-adjacent preference may be shaped less by static identity categories than by practice history, confidence, relational skill, and the ability to imagine implementation. Partnership-status and couple-configuration signals remain too sample-limited for strong inference.
A Typology of Preference Regimes
The central result can be summarized as five activity-response regimes that may coexist within the same person.
Low-Friction Partner-Contained Display
This regime includes sexy photos, private videos, partner masturbation while watched, and other forms of controlled erotic visibility. It is broad, dyadic, comparatively low-logistics, and less about generic exhibitionism than about being desired by a known partner, creating a private record, or making the partner an intentional witness.
Negotiation-Heavy Expansion
This regime includes group play, playing with another couple, swinging with multiple couples, mixed-gender group play, nude beach contexts, and online sexual posting. Elevated maybe responses suggest conditional openness: the desire may be real, but users need details about people, setting, privacy, safer sex, gender composition, boundaries, aftercare, and reversibility.
Niche High-Salience Relational Theater
This regime includes cuckolding and hotwife-adjacent scenarios where partner desire, witnessing, proximity, and primary-bond meaning are central. It is less broadly accepted than private exhibitionism but salient to motivated subsets, partly because the same surface can mean compersion, pride, eroticized jealousy, surrender, partner admiration, reclamation, threat, abandonment, comparison, or loss of status.
Linguistic Humiliation and Comparison
This regime includes harsh cuckolding phrase content. It is strongly rejection-heavy overall but can be highly salient among those who want it. Words create the frame: a phrase can transform an outside-partner fantasy into degradation, masculinity comparison, erotic shame, emotional masochism, or dominance.
Sparse or Undermeasured Voyeurism/Exhibitionism
The Voyeurism or Exhibitionism subcategory has high apparent acceptance but very low user coverage in the analyzed report. It should be treated as a signal that needs more data, not as a confident regime estimate. Future analyses should separate invited watching, partner-mediated watching, public audience play, digital exposure, and spontaneous/nonconsensual frames.
Discussion
The analyzed data do not support a one-appetite model of non-monogamy-adjacent kink. They support a multiple-regime model in which activities differ by privacy, audience, logistics, partner involvement, emotional tone, language, reversibility, and relational meaning. Treating those responses as one average score discards the main structure.
For practicing psychologists and sex therapists, the key clinical point is that labels are too coarse for assessment. "Cuckolding," "hotwife," "swinging," "voyeurism," and "exhibitionism" should be followed by concrete questions:
- Who is present?
- Who is watching?
- Who is allowed to touch whom?
- Is this fantasy-only, online-only, private, public, or in-person?
- Is humiliation desired, forbidden, or uncertain?
- Does the primary partner participate, witness, direct, reclaim, or stay away?
- What happens before and after?
- What would make this a yes, a maybe, or a no?
These details are not therapeutic over-specification; they are where preference lives. Consent education should likewise distinguish acceptance from salience, curiosity from readiness, fantasy from enactment, and erotic language from relational meaning.
Implications
Research Implications
Future analysis should model non-monogamy-adjacent preference as multidimensional, using latent class analysis, item-response models, or multilevel models that separate user, activity, role frame, subcategory, and wording. Maybe responses deserve their own model because collapsing maybe into yes-or-no loses one of the clearest signals in the data. Head-to-head salience should also be analyzed separately from swipe acceptance. Experience level deserves longitudinal study because the current analysis cannot distinguish learning, comfort, selection, or tenure.
Clinical and Educational Implications
The data support a structured conversation model with at least four columns:
| Column | Meaning |
|---|---|
| Yes | I want this or am ready to try it under ordinary conditions. |
| Maybe | I need more detail, trust, timing, structure, or negotiation. |
| No | I do not want this. |
| Salient fantasy | This pulls attention even if it is not ready for enactment. |
That fourth column matters. A fantasy can be salient and still not be safe, desired, ethical, or timely to enact; an activity can also be broadly acceptable but not especially important.
Practitioners should also separate activity, context, language, and aftermath. For example:
- Activity: partner has sex with someone else.
- Context: fantasy-only, same room, different room, supervised, club, online, known person, unknown person.
- Language: pride, compersion, humiliation, comparison, tenderness, ownership, neutrality.
- Aftermath: reclamation sex, cuddling, debrief, privacy, disclosure, no details, full details.
Each layer can flip preference.
Product and Community Implications
Preference systems should not recommend non-monogamy content as if it were one ladder. A better system would distinguish:
- Broad green-light items: high yes and high yes-or-maybe.
- Negotiation items: high maybe.
- Polarizing items: balanced yes and no.
- Niche high-salience items: lower broad yes but strong head-to-head performance.
- Strong no items: high rejection, especially where wording is harsh or identity-loaded.
Community education can use the same distinction. The goal is not to move everyone toward more intense content, but to identify the conditions under which desire, consent, safety, and relational meaning align.
Limitations
The dataset is not representative of the general population. It reflects engaged BeMoreKinky users interacting with kink and relationship content inside the app. Heavy premium gating means access, willingness to pay, and app tenure are structural confounds. Demographic fields include large unknown buckets, some subgroups are sparse, and the Voyeurism or Exhibitionism subcategory is especially sparse.
App behavior is not enacted sexual behavior. A yes response may mean fantasy, willingness, prior experience, curiosity, or readiness. A no response may mean permanent refusal, current relational mismatch, insufficient context, or dislike of the wording. The strength of the data is the large-scale patterning of preference disclosure inside a consent-and-communication app.
Conclusion
The analyzed BeMoreKinky data show that non-monogamy-adjacent kink is not one appetite. It fragments into broad partner-contained display, negotiation-heavy group curiosity, niche but salient cuckolding and hotwife-adjacent relational theater, rejection-heavy humiliation language, and undermeasured voyeurism/exhibitionism. That structure aligns with CNM literature showing subtype-specific stigma and perception rather than a single CNM construct (Matsick et al., 2014; Balzarini et al., 2018) and with fantasy research showing that arousal, discomfort, desire, and enactment should be separated analytically (Joyal et al., 2015; Joyal & Carpentier, 2017; Canivet et al., 2022).
The most important practical finding is that yes, maybe, no, and chosen-over-another-option should not be collapsed. Yes marks current acceptance; maybe often marks missing relational infrastructure; no marks a boundary; forced-choice salience marks erotic pull. This distinction is consistent with evidence that desire, familiarity, and engagement separate in polyamory (Moors et al., 2021) and that CNM fantasies can be salient without being ready for enactment (Lehmiller, 2020).
The clinical and product implication is item-level specificity. Practitioners and apps should separate activity, context, audience, language, risk, and aftercare, especially around group play, public/digital display, and cuckolding scripts. Consent and compersion research show that framing changes meaning: the same partner-with-other scenario can become threat, pride, compersion, humiliation, or ambivalence depending on agreement and script (Mogilski et al., 2019; Flicker et al., 2021). BDSM consent frameworks then explain why context-sensitive negotiation is required before moving from interest to practice (Williams et al., 2014; Tarleton et al., 2025). For users, the useful permission is simple: wanting one version of non-monogamy does not require wanting all versions, and rejecting one script does not invalidate another desire.
The practical companion to this report: Non-Monogamy Is Not One Appetite