The argument
Two years ago this was a thesis. Today it's a business: $873K gross, 2,615 students, and a base that settled at $35–50K a month rather than falling to zero. It also taught us three things we could not have known going in. 85% came for access and we charged for the course. Four attempts at upsells returned a fifth of plan. And it took five months and a 76% price cut to find the number that worked. The next platform can start with all three already solved.
Belonging first — the room, the creator, the people. 85% joined for that. Then a real craft: practice with feedback from working professionals, and a credible path into an industry that is otherwise full of noise and scams.
We build and run all of it — course, site, funnel, paid media, community, events, payments, support — at no cost to them. They get a seat in our ops Slack and work beside the team rather than through an account manager. They keep ownership and final say. Tara's own audience was worth about $250K. The machine turned it into $873K. That multiple is what the share buys.
| Today | 2029 | |
|---|---|---|
| What we sell | A course, bought once | Access, renewed every year |
| Revenue shape | Resets when the audience runs out | Compounds platform on platform |
| How we build | Bespoke every time | Modular sections we snap together — swap the narrative without a rebuild |
| Who we sign | Warm intros and instinct | A revenue gate, applied before the meeting |
| What we own | A revenue share | Equity with a buyout price |
| Year three | Undefined — and we're there now | Written into the deal before we sign it |
The whole bar is what we do today. Thirty points are named, twenty are still to find — most likely in a component library so builds stop being bespoke, automated reporting, retiring guest module production, a local crew for course shoots, and folding expansion packs and small groups into the membership instead of marketing them separately.
| Function | Class | Why |
|---|---|---|
| The engine — every platform, always | ||
| The demand system Paid ads · email · organic social · SEO and blogs · influencers · affiliate · community engagement | Lion | Not seven channels. One ecosystem, and together it is the 10x. Paid takes the attributed credit because it is the only one that reports on itself — but it is harvesting demand the others create. Community engagement is the loop closing: members who feel it tell people, and that comes back as demand we never paid for. Silo any of them and the number moves. |
| Course, site, checkout | Lion | Everything in the system drives traffic to one place, and this is it. Only 11% joined for the modules, but the course is still the thing being bought — it's the transaction the whole machine exists to produce. |
| Live Zooms, AMAs, gyms | Lion | This is where our content comes from. No live events, no footage, no creative for the 10x. Only 3% joined for them, but they feed the machine and members name them constantly. |
| Modular site system | Lion | Sections that snap together — hero, proof, offer, pricing, objections, FAQ. We test a new narrative or marketing angle in an afternoon instead of a sprint, and platform #2 gets assembled rather than built. This is what makes the tenth one cheap. |
| Free, and it's why they join | ||
| Community — stays on Facebook | Lion | 54% joined for it, 66% stay for it. Meta killed the Groups API, so it can't be gated — and shouldn't be. |
| Contests | Lion | Part of the 31%. Acquisition, plus the success stories that feed ads. |
| The renewal — what we start charging for | ||
| 1:1s with the creator | Horse | The rest of that 31%, and today it's only a raffle prize. Most-requested thing we don't sell. |
| Small-group coaching | Horse | Four emails and one creative asset isn't a test. Members ask for practice and feedback constantly. |
| Theirs, not ours | ||
| Organic posting | Their lion | Their reach drives the spikes. We produce every asset — but they have to post it, and today nothing makes them. That gap is the single biggest unfixed risk in the model. |
| Fix it, don't cut it | ||
| Posting Zooms to the portal | Dog today | Recordings sit for weeks, supplement PDFs get missed. Fulfilment problem, not a scope one. |
| Stop | ||
| Commercial features | Dog | Roughly 20% of company bandwidth for under 5% of impact. The clearest misallocation we have, and the easiest thing on this list to stop. |
| Text threads with the creator | Dog | Requests arrive at all hours, get lost, and nobody else can pick them up. One channel, agreed response times, and a standing weekly call for everything that isn't on fire. A seat in our ops Slack becomes part of what we sell, not an afterthought. |
| Guest modules | Dog? | Members name the guest teachers constantly — but always in the live sessions. Recorded guest modules look like the same pattern as our own: the access lands, the artifact doesn't. Confirm before we cut. |
| Studio-recorded course production | Dog | This is how signed deals die. Music Travel Love never finished recording. Gab exited frustrated with recording and editing. The mistake is handing it back — we run everything else, then ask the creator to shoot and edit. Fix: hire a local crew in their market, one shoot day, fast-turnaround edit. Their obligation drops from thirty hours and an edit suite to a single day, and live sessions cover the rest. A day of local production costs a rounding error against a $500K platform. |
| Merch, job board, store | Dog | Survey interest was high, actual sales are negligible. Products that never generated demand. |
Teaching intent is the criterion we were missing. "Not looking to teach" is what kills deals, and no creator has ever countered on 50/50 — the split was never the obstacle. Target people already in motion: an announced course, a waitlist, or one that launched and went quiet. Then Ameet's waitlist test — 1,000 qualified leads before we build anything.
| Vertical | Buyer | Why this one |
|---|---|---|
| Medical aesthetics | Injectors, med-spa owners | Biggest ecosystem of the four, and the buyer expenses it — a credential that raises a billable rate. Flag: Meta classifies med-spa domains under health and wellness and throttles conversion data, so we test a domain before we commit. |
| Physical therapy | Clinicians | Continuing education is already a professional obligation. Annual renewal is normal here rather than novel — the hardest part of our model is pre-sold. |
| Culinary arts | Cooks, serious enthusiasts | Endlessly visual, so the content engine feeds itself. Consumer buyer though — lower ticket, needs scale to clear the gate. |
| Golf | Amateur players | The most provable outcome of any vertical here — handicap is a tracked, public number. Affluent by default, enormous existing spend on instruction, and a swing is about as AI-proof as a skill gets. |
| Tattoo | Working artists, apprentices | No formal training path and an apprenticeship bottleneck, so the demand has nowhere else to go. Strongest artist-level fandom of the four, and a portfolio is proof you finished. |
All four share the three things voice acting couldn't give us: a craft you can't learn from a machine, a buyer with money, and an outcome you can point at.
The table above — core, renewal, creator's, stopped · creator posts under a content SLA · events on one shared operator · anything new gets a launch attached or it doesn't ship
$500K projected year one · professional verticals priced at $2,500–5,000 · teaching intent and the waitlist test as hard gates
The renewal is access to the creator — 1:1s, small groups, critique, live sessions · course owned for life · community never expires · first year folded into the price · Voicelings stays as sold
Every agreement states what year three is · a buyout at a set multiple of trailing revenue, or an agreed wind-down · no platform signed without one
Two things I haven't answered and would rather we argue than assume. Do we build platforms from scratch, or take on creators who already have a course that stalled? At one or two signings a year that choice decides whether ten by 2029 is arithmetic or ambition — but the second path carries a risk worth naming: if their audience already bought once, the buyers are spent, and our whole model depends on an untapped launch clearing ~$100K before paid can start. And is the 40% equity vesting after $100K already agreed? It's the only equity language any of us has written down, and decision 04 assumes it's still open.
Appendix below — the model you can drive yourself, every number on this page with its source and confidence, and the build spec for the membership.
Three things: how the membership actually gets built, the model behind the gate, and where every number came from. Open whichever you want to push on.
Stage one runs on the creator's audience alone, because we put in no cash. It has to clear ~$100K before there is anything to fund an ad account with. Stage two only exists once it does. Note what the model doesn't have: a size input for our effort. The build costs the same whether the creator has 100K followers or two million — same course, same site, same funnel, same team. That is why the gate is about what comes out, not what goes in.
Two engines, and you can see the handover. Everything gold and navy is Tara's own audience — the same people, priced wrong for five months and then right for one. That single month is what her reach was worth: about $223K all in. Everything violet after it is paid, and it's the part that didn't decay — roughly $500K a year at platform level, still running, still rising.
Two are load-bearing and unconfirmed. They are marked, and the model is only as good as they are.
| Figure | Source | Sample | Confidence |
|---|---|---|---|
| Why members joined | Voicelings group poll, "What made you want to join the most?" — private group with Tara 54%, contests and 1:1s with Tara 31%, modules 11%, live events 3%, script library 1% | 220 votes, 409 reach · Nov 2025 | Confirmed |
| 85% joined for access | Derived: private group 54% plus contests and 1:1s 31% | Same poll | Confirmed |
| 66% name the community, 13% name Tara | Member feedback survey, open text, theme-coded | 140 responses · Jan–Feb 2026 | Confirmed |
| $870K, 2,615 students | Partner strategy docs, two-year cumulative | 2024–2026 | Confirmed |
| $200K planned, ~$40K real | 3.0 module expansions and facilitated practice. Ameet's summary states $200K+ as delivered; the source section states it as projected | — | Needs reconciling |
| $5K in, $50K out — 10x | Ad account. Implies ~$29 a sale on a $290 seat, and lead-to-sale near 10% | July 2026 | Confirmed |
| ~1% conversion | Site visitor to purchase — a different denominator from lead to sale. Do not divide cost per lead by it | Unverified | Jaro to confirm |
| 10% reach per post | Calibrated to Tara: 1.5M audience, 3–5 posts, about 700K views | July 2026 | Estimate, locked |
| Merch demand | Member feedback survey: 4.1 items and a $156 basket on average, everyone picked something. Actual sales are negligible | 140 responses | Contradicted by sales |
| $500K gate | Proposed here. Not observed — it is the decision, not the evidence | — | For Monday |
Two plans that stack. Memberstack and Stripe handle the state change between them, and no human ever removes anyone from anything.
One row pays once. The other pays every year. Year one is folded into the purchase price, so the first renewal lands twelve months after they join — and renewals stagger naturally by join date rather than all landing in the same month.