Skip to content

Customer Voice, Demand Signals & Audience — Raw Findings

Phase: 3 — Wave 3 (consolidated, C1+C2+C3+C4) Date: 2026-05-09 Confidence: Medium-High on macro signals; LOW on Likeness-specific creator voice (no primary research has been done; this is the largest validation gap)

Customer Voice (C1)

What we know from public discourse

Public sentiment on adult deepfakes (Tier 2 academic): - Reddit-wide deepfake sentiment study (17,720 posts/comments): 47.0% positive, 36.8% negative overall. - Adult deepfake abuse specifically: 47.5% negative — the highest negative score of any deepfake topic. [Springer / Tier 2] - Strongest moral intuitions in negative-sentiment posts: Sanctity/Degradation. Strongest in positive-sentiment posts: Care/Harm (community support). - AI-IBSA (image-based sexual abuse) targeting intimate partners is viewed as more unacceptable than targeting celebrities. [Tandfonline / Tier 2]

Implication for Likeness: The cultural license for "consent-first" framing is strong. Public attitudes are moving against unauthorized AI use. The platform's narrative aligns with this trajectory — it's the legitimate alternative to a thing the public increasingly opposes.

What we don't know — the discovery gap

The founder Q&A confirmed: - Zero structured creator interviews have been conducted. - The reviewer-checklist.md document exists but has not been used. - The mockup is shareable but has not been shared with target creators.

This means we have no primary data on:

  1. Whether adult creators actively want a platform like Likeness.
  2. What features rank highest in creator priority order.
  3. How creators feel about the verification + license-gating + no-export commitments.
  4. Whether the proposed economics (20% platform take + AI compute markup) feel acceptable.
  5. Whether the mockup's voice and design feel respectful or condescending.
  6. What concerns creators have that Likeness's planning has not anticipated.

This is the largest single validation gap in the project. Filling it is Validation Phase priority #1.

Indirect creator signals (Tier 2/3)

From industry coverage and platform CEO interviews: - "Real-life adult-content creators have concerns about AI affecting their business — some find AI tools useful (chatbots for fan interaction) but worry that using AI could erode fans' trust." [Yahoo / Tier 2] - OnlyFans CEO Keily Blair has spoken cautiously about AI's influence; she came from "cyber, privacy, and online safety" background. [Financial Times / Tier 2] - 12,000+ deepfake-violation account bans on OnlyFans 2024-2025 indicate active enforcement and creator complaints.

Implication for Likeness: Trust erosion is the named concern. Likeness's positioning ("you're in charge of every decision") directly addresses this — but the addressing is theoretical until it's tested with a creator who can say "yes, this version of AI on my platform feels right; this version doesn't."

Demand Signals (C2)

Quantified demand for adult AI content

  • Adult AI image generation market: $2.8B in 2025 (+340% from $820M in 2024) [Tier 3 — directional]
  • AI companion apps revenue: $120M projected 2025 [Tier 2]
  • Combined emotional + NSFW AI companion segment: $1.2B in 2025 at 32% CAGR [Tier 3]
  • 337+ active revenue-generating AI companion apps as of July 2025; 128 new launched in 6 months [Tier 3]

Quantified demand for creator-side AI on legitimate platforms

  • 15% of Fanvue revenue is AI-generated content [Sacra, Tier 2] — the strongest single signal that fans WILL pay for AI on creator subscription platforms.
  • OnlyFans 2024: $7.22B GMV, 4.634M creators, 377M fans [Variety, Tier 2].

Pricing intelligence

Reference point Price Source
OnlyFans subscription range (typical) $4.99-$49.99/month Tier 3
OnlyFans recommended starting price $9.99-$12.99/month Tier 3
OnlyFans top-tier creator subscriptions $50-$100+/month Tier 3
Adult AI generator subscription (Promptchan) $5.99-$18.99/month Tier 3
Adult AI companion (Candy.ai) $5.99 annual / $12.99 monthly + $0.40/image Tier 3
OnlyFans % from one-off sales (PPV, customs) 59% vs. 41% subscriptions Tier 3

Key insight: The bulk of OnlyFans creator revenue comes from one-off sales, not subscriptions. Likeness's revenue model needs to support PPV / submission fees / per-generation pricing as first-class flows, not just subscriptions. The founder brief gets this right (compute credits + submission fees + custom requests) but the relative weighting matters.

Implications for Likeness pricing tiers (from founder brief)

The illustrative subscriber tiers in the founder brief ($15 / $25 / $50 / $100 / $200) are calibrated against existing OnlyFans tier ranges and are reasonable. However, given that 59% of revenue is one-off in the comparable market, the tier structure should be pitched as floor pricing, with most creator revenue coming from per-generation and submission fees on top.

[Estimate] Realistic per-creator monthly revenue distribution at maturity: - Subscription floor: $15-$50/month per fan, ~10-30% of GMV - AI generation credits: $20-$200/month per active fan, ~30-40% of GMV - Submission fees + custom approvals: $0-$50/month per fan, ~20-30% of GMV - PPV unlocks of approved generations: ~10-20% of GMV

This shape is closer to OnlyFans economics than to a pure SaaS-style consumer subscription, which is correct.

Audience Profiling (C3)

Persona 1: Mid-tier adult creator with existing audience

Profile: - Age 22-40, US-based for MVP - Operates on OnlyFans / Fansly / direct platforms; possibly multiple - Existing audience: 1K-50K active fans / subscribers - Monthly gross revenue: $5K-$80K (the meaningful tier — top 1-10% of OnlyFans creators is roughly this band, derived from $49K/year top-1% figure and known long-tail distribution) - Tech-savvy enough to manage multiple platforms; not necessarily an AI expert

Pain hierarchy (synthesized, NOT validated with primary research): 1. Unauthorized AI use of likeness — happening already, no recourse, no revenue, fan confusion. 2. Revenue ceiling on real content — there are only so many hours in the day; AI is a potential additive revenue line. 3. Platform risk — Mastercard/Visa rule changes can change earnings overnight; some categories of content get banned. 4. Trust with fans — fans care about authenticity; bad AI use erodes trust. 5. Compliance overhead — 2257, ID verification, takedowns are real time-cost.

[Critical caveat]: This pain hierarchy is logically derived, not surveyed. The order is plausible. The actual order, especially the relative ranking of #1 vs. #2, is an open question that creator interviews would resolve.

Buying behavior: - Decision-making is individual (creator chooses their own stack), but agency-influenced for higher-tier creators. - Trust is the gating factor. Creators are skeptical of new platforms. - Adoption motion: try with a small cohort, expand if it works. - Time to first revenue is critical — concierge-phase creators need to see money fast or they churn.

Where to reach them: - Industry events (XBIZ, AVN) - Creator-managed agency relationships - Existing Discord and Telegram creator-community channels - Founder-led 1:1 outreach (the de-facto playbook in adult industry)

Persona 2: Engaged fan/subscriber

Profile: - Existing OnlyFans / Fansly subscriber - Pays $20-$200/month across multiple creators - Some willingness to pay for "extras" (PPV, customs, AI) - Fan/creator parasocial relationship is a real driver

Why they'd use Likeness: the AI version of their favorite creator gives them content the creator can't or doesn't want to produce manually. Per-generation pricing fits a "tip economy" mental model.

Risks for the fan side: - Fragmented payment relationship (one more platform to subscribe to). - Friction of moving from a known creator account to Likeness. - Compute credit pricing must feel fair vs. the volume of generation.

Pricing Deep-Dive (C4)

The C2 section above covers most of the pricing data. Three additional points:

  1. Adult-friendly processor fees compress margins. With CCBill / Segpay at 5-15% (vs. 2-3% mainstream), Likeness has 5-10 fewer margin points than a generic SaaS at every tier. This is a structural cost of operating in the category.

  2. Token-based generation pricing is the standard. Candy.ai $0.40/image, Promptchan tiered limits. Likeness's "compute credits" model lines up with this.

  3. Anchor pricing for the AI generation tier should reference adult-AI-generator pricing, not OnlyFans subscriptions. A fan considering paying $50/month for "real content + AI generation credits" (founder brief's tier 3) is mentally comparing against (a) the existing OnlyFans tier, AND (b) the existing AI girlfriend subscription. Both anchors must be cleared.

Flags

Red Flags: - Zero primary creator data is the load-bearing risk. Every product, pricing, and positioning decision in this skill's output rests on inferred creator preferences. If primary discovery reveals different priorities or unanticipated objections, much of this work has to be revisited. This is named in Phase 8 (Validation) as the highest-priority experiment.

Yellow Flags: - The PPV / one-off sale weighting (59% of OnlyFans creator revenue) means Likeness's pricing logic must support a wide range of one-off transactions cleanly. Friction in those flows kills creator earnings. - AI generation as a fan tier may or may not be incremental to existing fan spend. If it cannibalizes real-content spend (i.e., fans spend $50 on AI instead of $50 on real content), creator total earnings don't increase — and creator buy-in fails.

Sources