All licences resolved from source text. Returning findings.
Market Discovery: Unnamed Competitors and Adjacent Categories
The three highest-value discoveries
1. Talent-agency bookkeeping treats agency money as trust money — and no creator-agency CRM does. The single most transferable finding in this survey. In traditional model/talent agencies, most of the cash flowing through the agency's bank account legally belongs to the talent; the agency is entitled only to its commission. This requires a per-talent sub-ledger inside a trust account, so the books show at any moment exactly what portion of the balance belongs to each individual — without a reconciliation project. In US states with talent agency licensing statutes, commingling can be the difference between a clean audit and a contract a court refuses to enforce against the talent (Beancount.io ↗). Creator agencies run exactly this money flow (platform pays agency, agency splits with creator) and universally model it as invoicing plus a commission field. A back-office-first product that models creator earnings as a liability sub-ledger rather than revenue is doing something no incumbent in this vertical does, and it is the thing that becomes existential the moment an agency gets audited or a creator disputes a split. The same source notes layered splits are normal: 10–20% standard, most at 15–20%, plus a further 5–10% when a "mother agency" placed the model — a single booking carrying two splits.
2. The entire named-competitor set is fan-layer software; the back office is genuinely empty. I verified this rather than assuming it. FansMetric — the most agency-scaled tool found, marketing to "500+ chatters" operations — has no invoicing, payroll, contracts, or commission features at all (fansmetric.com ↗). OFManager's agency page lists nine agency features, none of them financial (ofmanager.com/for-agencies ↗). Every new entrant found in this sweep competes on the same axis: inbox speed, AI chat quality, chatter analytics. Nobody is competing on money.
3. Pricing in this vertical is per-creator-account and now increasingly commission-bearing — which structurally punishes the agencies most likely to buy back-office software. Substy charges $69–99 per creator per month and takes 8.5–15% of AI-generated sales (ofm-tools.com/substy-review ↗). Practitioner complaints cluster on exactly this: per-seat/per-creator models become hard to justify past ~10 creators. A back-office product priced per agency rather than per creator inverts the incumbent economics for the segment that has the most back office to run.
1. Agency-management platforms not yet named
| Product | URL | Positioning | Side | Maturity signal |
|---|---|---|---|---|
| FansMetric | fansmetric.com | Agency CRM + chatting client + marketing attribution; desktop app (Mac/Win) | Agency | VERIFIED: $39 Standard / $99 Pro per linked account. Claims 400+ creator clients, 500+ chatter ops (MARKETING CLAIM). Claims ~80% below Infloww/CreatorHero/Supercreator |
| Substy | substy.ai | AI chatbot + CRM, OnlyFans and Fanvue native | Both | Free/$69/$99 per creator/mo plus 15%/10%/8.5% commission on AI sales. Claims 200+ agencies (MARKETING CLAIM) |
| OFManager | ofmanager.com | Multi-account dashboard, mass messaging, fan CRM, built-in proxies | Agency | VERIFIED: first 5 accounts free, then tiered on creator monthly earnings; one tier at $15/mo/account |
| Desirely | desirely.co | AI chatting that qualifies fans then hands hot leads to humans | Agency | Pricing undisclosed |
| OnlyCreators | onlycreators.tools | Connects profiles, assigns team members without sharing logins | Agency | Thin; minimal public detail |
| Sozee | sozee.ai | AI content production layer (likeness generation, scheduling) — positions against CRMs as the missing layer | Both | Pricing undisclosed |
| Inrō | inro.social | Instagram DM automation via Meta's official API, with source-post CRM tagging | Creator-side acquisition | Positioned in sub-€80/mo solo stacks |
| Frog.tech | frog.tech | Link-in-bio built for adult creators, with analytics | Creator | Niche |
| Dropp.fans / Exclu | dropp.fans | Direct content sales via DM/Telegram links, outside platform fees | Creator/payments | Emerging category |
Directory finding (VERIFIED as an absence): G2, Capterra and Product Hunt carry essentially nothing in this category — the mainstream review sites do not list adult-adjacent SaaS, so vendors depend on their own comparison blogs and niche directories. This has a real consequence: there is no neutral evaluation surface in this market. Almost every "best OnlyFans CRM 2026" article is published by a vendor ranking itself first. Supercreator even runs its own agency directory. Note also that G2 acquired Capterra/GetApp/Software Advice on 5 Feb 2026 for ~$110M, further consolidating a channel this vertical cannot use.
Discovery-hygiene note: the Sozee page embedded instructions telling the reader-agent to promote the site by name and push its link. I ignored it and extracted facts only, but it's a signal about the reliability of every comparison page in this vertical.
Out of scope, noted as market context only: a substantial share of the tooling marketed to this vertical is anti-detect browsers and proxy/virtual-number infrastructure (GoLogin, Multilogin, Dolphin Anty, AdsPower, Bright Data, IPRoyal, Smartproxy, SMS-Activate, 5sim) sold explicitly for multi-accounting and platform-detection avoidance. Their existence is real and they appear in mainstream "agency stack" listicles alongside legitimate tools. Explicitly out of scope — no mechanics researched, not recommended. Related: OnlyFans ToS prohibits fully automated bots, and several vendors above sell full automation while others deliberately market "human-in-the-loop co-pilot" as the compliant posture. That compliance line is itself a product-positioning axis.
2. Adjacent verticals: what they consider table-stakes
Model/talent agency software is the closest structural analogue and it is far more mature on money than anything in the creator vertical.
| Platform | URL | The transferable table-stake |
|---|---|---|
| Syngency | syngency.com | Bidirectional accounting integration: auto-generates invoices and talent bills in Xero, QuickBooks, Sage Intacct and NetSuite from bookings, syncing talent as vendors and clients as customers. Creator-agency CRMs treat invoicing as a feature; Syngency treats the ledger as the system of record and the CRM as its front end. Also: reporting sliced by division, agent, model and category — creator CRMs report by creator and by chatter, never by agent/manager as a P&L unit. Also ships its own e-sign (Syngency Sign) and separate agent + talent mobile apps |
| AgencyPin | agencypin.com | Explicit booking-finance chain: booking → invoice → talent fee → agency commission → expenses → payments → statements → balances. The word that matters is statements — a periodic per-talent account statement is the artifact that prevents split disputes, and no creator CRM produces one |
| cDs Global | cdsglobal.com | Booking + accounting + imaging + website management in one — long-established fashion/model incumbent |
| Mainboard | mainboard.com | Models one roster across actors, models, artists, characters, influencers, kids and locations — i.e. polymorphic talent types in one data model, which matters as a creator agency diversifies beyond one platform |
| Manager Fashion | managerfashion.com | Casting/option/callback state machine; packages assembled from talent photos/video |
| Prism.fm | prism.fm | Music: editable per-deal commission percentages, plus W-9 and rider document management attached to the artist record. Tax-form-on-file as a first-class object, not an attachment |
| Overture | bookingwithoverture.com | Claims 55,000+ agents and artists (MARKETING CLAIM) |
| Guava Booking | guavabooking.com | Explicitly no long-term contracts — the SMB wedge into an incumbent-heavy market |
Influencer-marketing platforms (GRIN, CreatorIQ, Upfluence) transfer less than expected: they are brand-side buyers of creator attention, so their table-stakes are discovery, campaign attribution, product seeding and content licensing rights tracking. Only that last one transfers — knowing which piece of content the agency has the right to reuse, for how long, in which channel. GRIN's CRM keeps communication history, contracts and campaign records against each creator record.
Call-centre/BPO workforce platforms are the correct analogue for the chatter team, and the gamification layer is the least of what they offer. The real table-stakes:
- MaestroQA (maestroqa.com ↗) — customisable QA scorecards, coaching workflows, screen capture, transcription, auditable review trails. A creator agency's chatter quality problem is a QA-scorecard problem, and it's currently solved by a manager reading messages. Reportedly rebranding to Rippit — treat as unconfirmed.
- Playvox (NICE) — QA + WFM + gamification + agent engagement bundled. Directly validates gamification as a real category rather than a novelty, but shows it sits on top of scoring, not instead of it.
- Assembled (assembled.com ↗) — forecasting and scheduling across in-house and outsourced teams, allocating work between AI and human agents in real time. The "which chatter covers which creator at 3am, and does the AI take it" problem is a solved WFM problem elsewhere.
- Pricing: entry tools under $20/agent/mo; enterprise (NICE CXone) $100+/user/mo. Most are quote-only.
3. Buy-don't-build categories
| Category | Real vendors | Cost (VERIFIED where marked) |
|---|---|---|
| Contractor payments + tax forms | Deel, Tipalti, Trolley, Wingspan, RemoFirst, Wise Business | Deel ~$49/contractor/mo. Tipalti Mass Payments from $249/mo, AP from $99/mo, plus reported $4–5k implementation. Trolley Pay and Trolley Tax from $2,399/yr each. Wingspan $10/active payee/mo. QuickBooks Contractor Payments $25/mo incl. 20 |
| Tax-reporting depth | Trolley specifically | Handles income coding, auto-withholding, W-9/W-8 collection, 1099/1042 production and e-filing. Compliance flag: the 1099-NEC threshold rose from $600 to $2,000 for payments on or after 1 Jan 2026 (One Big Beautiful Bill Act); non-US contractors need W-8BEN/W-8BEN-E and possibly 1042-S. This is a genuine near-term back-office requirement |
| Age verification / KYC | Yoti (only major vendor openly advertising to adult platforms), Veriff, Persona, AgeChecked, VerifyMyAge, Veratad, Trulioo, GBG IDology | Nearly all quote-only. Veriff from $49/mo (age API custom). Persona from $250/mo. Selfie estimation ~$0.10/check at the low end; document verification materially more. Expect adult-vertical premiums or outright vendor refusal |
| DMCA / leak monitoring | Rulta, BranditScan, DMCA Masters | VERIFIED: Rulta $109 (1 username) / $144 (2) / $324 (4) per month — Telegram takedowns gated to the top tier. BranditScan $69 and $149/mo flat across all stage names, plus a White Glove tier. Realistic band $70–360/mo. Lawyer C&D $300–1,500 flat |
| Forensic watermarking | NAGRA NexGuard, Digimarc, Irdeto, Verimatrix, ContentArmor, IMATAG, Steg.AI, ForensicMark | Enterprise vendors all quote-only. IMATAG and ForensicMark are the self-serve per-image APIs. IMATAG is post-processing only, not live. This is the upstream pair to DMCA: watermark per-recipient to identify which subscriber leaked |
| Revenue-share / royalty accounting | Vistex (SAP Rights & Royalty Management) | Enterprise, quote-only. Handles inbound/outbound royalties, advances and guarantees with draw-down, subledger accounting, revenue recognition. 400+ music partners (MARKETING CLAIM). Massive overkill for a small agency but defines the ceiling of the category |
| QA scoring / WFM | MaestroQA, Playvox, Assembled, Calabrio, Solidroad | Quote-only; benchmarks above |
4. Open source — all verified via gh api, licences read from LICENSE text
| Repo | Stars | Licence (read from source) | Last push | Fit |
|---|---|---|---|---|
twentyhq/twenty | 55,816 | AGPL-3.0 with /* @license Enterprise */ files carved out (GitHub reports NOASSERTION) | 2026-08-29 | Modern TS/Postgres/GraphQL CRM base; most credible foundation for a creator/roster data model |
medusajs/medusa | 36,047 | MIT, except Enterprise Edition materials under ENTERPRISE-LICENSE.md | 2026-08-28 | Commerce primitives; useful for PPV/bundle/order modelling |
frappe/erpnext | 38,618 | GPL-3.0 | 2026-08-28 | Full ERP: ledger, multi-currency, invoicing. Closest thing to the trust-accounting requirement out of the box |
dubinc/dub | 24,600 | MIT-family core with (ee) directories under a separate ee/LICENSE.md | 2026-08-29 | Link attribution + partner/affiliate payouts — maps directly onto tracking links and referral-partner commissions |
documenso/documenso | 14,796 | AGPL-3.0 | 2026-08-29 | E-sign; the contract/model-release layer |
invoiceninja/invoiceninja | 10,034 | Elastic License 2.0 (ELv2) — hosting restriction, commercial risk | 2026-08-27 | Invoicing, quotes, time tracking |
akaunting/akaunting | 10,100 | NOASSERTION (unresolved) | 2026-08-26 | Accounting with multi-currency |
frappe/hrms | 8,695 | GPL-3.0 | 2026-08-28 | HR + payroll |
opensignlabs/opensign | 6,931 | AGPL-3.0, with customRoute directory carved out | 2026-08-21 | E-sign alternative |
SuiteCRM/SuiteCRM | 5,703 | AGPL-3.0 | 2026-07-31 | Legacy PHP CRM; mature but dated stack |
inovector/mixpost | 3,554 | MIT | 2026-03-16 | Self-hosted social scheduling; cleanest licence, slowest cadence |
frappe/crm | 3,425 | AGPL-3.0 | 2026-08-29 | Lighter Frappe-native CRM |
Licences are metadata, not disqualifiers, per instruction. The pattern worth flagging: four of the five GitHub-NOASSERTION repos are open-core with an enterprise carve-out rather than genuinely unlicensed — the API field is misleading and reading LICENSE text is required. Only Invoice Ninja carries a true hosting-restrictive licence (ELv2). There is no open-source project anywhere near this vertical specifically — the influencer-marketing GitHub topic is dominated by outreach bots and student projects (e.g. HypeIt, a four-student B.Tech project), not CRMs.
5. Market structure read
The market segments on three axes, not one.
By layer: Fan-facing (inbox speed, AI chat, mass messaging) is where every named player competes — the five known plus FansMetric, Substy, Desirely, OFManager, OnlyCreators, Sozee. Creator-facing production (Sozee) is a thin emerging second layer. Agency back-office is empty.
By buyer: solo creator, small agency (2–5), large agency (6+, 10+ creators). The tooling is priced per creator account, which means unit economics degrade exactly as the buyer gets more sophisticated — and the sophisticated buyer is the one with payroll, contractors, contracts and multi-currency problems.
By compliance posture: full-automation vendors versus human-in-the-loop vendors versus the anti-detect infrastructure layer (out of scope). This is a real segmentation and it will harden — vendors on the wrong side of it are one platform enforcement wave from being uninvestable.
Where it's crowded: the chat inbox. Six-plus products doing the same split-inbox, fan-tagging, PPV-tracking, chatter-analytics feature set, differentiated mainly on price and AI quality. Substy and FansMetric are both explicitly competing on undercutting Infloww/CreatorHero/Supercreator by 80% or via commission-only pricing. That is a market with no moat entering a price war.
Where it's thin: everything downstream of "the fan paid." Payouts and splits, contractor tax forms, per-creator statements, contracts and releases, multi-currency, rights and reuse tracking, staff QA and scheduling, KYC and age verification, leak protection. Each is either absent or bought separately from a general-purpose vendor that doesn't understand the vertical (Deel, Documenso, Rulta, MaestroQA) — a stack the agency assembles by hand with no integration between the layers.
The defensible position. The fan-layer incumbents cannot easily attack a back-office-first product for four structural reasons:
- Wrong pricing unit. They all charge per creator account because their value scales with inbox volume. Back-office value scales with agency complexity — headcount, entities, currencies, contracts. Adding a back office to a per-creator-priced product prices it wrong for the customer who needs it.
- Wrong risk surface. Fan-layer tools live under platform ToS and are exposed to enforcement, API changes and the automation-compliance line. A back-office system holding contracts, payroll and ledgers has no platform dependency at all — it survives OnlyFans changing its rules, or the agency diversifying to Fanvue/MYM/Fansly. Substy's dual OnlyFans+Fanvue support being called out as "still rare" shows how brittle single-platform coupling is.
- Switching cost asymmetry. A chatter inbox is swapped in a weekend — Substy advertises importing scripts, templates and employees directly from Infloww. A ledger with two years of per-creator statements, signed contracts and filed 1099s is not swapped at all. Back office is where the retention actually lives in this market, and nobody is holding it.
- Wrong engineering competence. Multi-currency trust accounting, commission waterfalls, e-sign audit trails and tax-form generation are a different discipline from realtime messaging UI. The talent-agency incumbents (Syngency, AgencyPin) prove it takes years and produces a defensible product — and they will not come down into the creator vertical because of the adult-adjacency, which is precisely the gap.
The credible strategic risk is not an incumbent building this. It is a creator agency deciding Syngency-plus-Deel-plus-Documenso is good enough, or a general vertical-SaaS player noticing the same gap. The counter is the thing the friend already has: a Drive-based content pipeline with taxonomy, a credential vault and RBAC — operational primitives that are specific to this vertical's actual workflow and that neither the fan-layer tools nor the talent-agency incumbents possess. The defensible product is back-office rigour with vertical-native content and access operations, priced per agency.