This page contains confidential investor materials. Enter the access code to continue.
How O/U+ turns invested capital into acquisition instead of overhead — a tiny human core, an agentic technology stack that does the work of a department, and Partners who carry the licensing, inventory, and audience. Senior operating horsepower, without senior-hire burn.
O/U+ is not a software company and isn't capitalized like one. The moat is the proven venue model, the fan relationship, and the owned first-party data — not code. So we buy or rent the technology, partner for the heavy lifting, and point an agentic stack at the work — invested capital funds the two things that compound, creative and acquisition, rather than payroll and infrastructure.
We build essentially one thing: the fan relationship and the owned data. Everything else is rented, partnered, or handed to agents.
| Function | Decision | How we do it |
|---|---|---|
| Fan relationship + first-party data | Build & Own | The only real moat. Owned CRM/warehouse and owned agent logic — never let a partner hold this. |
| The AI avatar (VBA) | Partner | Specialist build partner (Biz4Group) delivers it; O/U+ owns prompts, workflows, credentials & data contractually; MCP keeps the agent layer swappable. |
| Licensing & compliance rails | Partner | WagerWire's affiliate licenses + the CPA model keep O/U+ in the light regulatory tier, nationwide. |
| National reach / audience | Partner | Prediction markets as the 50-state front door; venues, teams & creators supply warm reach we didn't buy. |
| Ad-ops & media buying | Platform | Platform AI runs targeting, bidding & optimization. No ad-ops hire. |
| Creative, ops & lifecycle | Platform + Agent | AI tools generate creative at volume; agentic ops do the repetitive work under human approval gates. |
| Growth capabilities ×3 | Platform | Audience intelligence, automated GTM, and data monetization licensed separately — no single supplier grading its own performance. |
| The technology function | In-house | Led by Frank Filippi (tech development) and Jordan McCrary (AI development) — the stack, integrations, data & attribution. |
Agents do: research & enrichment, drafting at volume, runtime personalization, and plumbing between systems. Humans keep: strategy & judgment, approval gates on anything touching money or contracts, closing, and taste. Everything else is not a person.
The stack, the integrations, the data, and — above all — attribution are led from inside O/U+. Frank Filippi leads tech development; Jordan McCrary leads AI development. Build partners deliver against their specs, and the asset stays with O/U+.
An operations leader — day-to-day execution across venues, partners, and the member funnel. Recruited against the launch and funded in the raise; until then the founder carries it with the agentic stack underneath.
Four partners carry the pieces that would otherwise be a licensing regime, a balance-sheet risk, a state-by-state legal calendar, or an audience we'd have to buy. Each one removes a function O/U+ would have to hire for.
A specialist AI build partner (Biz4Group) delivers the Virtual Betting Ambassador — phased deliberately, QR/website first, native app later, so launch never waits on an app-store cycle. Three things hold it: contractual ownership (prompts, workflows, credentials, data assign to O/U+), open connectivity (MCP keeps the agent layer swappable, not welded in), and a counterparty who can grade it — tech development leadership, on our payroll, from Day 1. If the build slips, the free-to-play front door and venue motion still generate onboards on a degraded agent.
Outsourcing the machinery is efficient but creates dependency. Prediction-market legal status is actively contested (a counsel-dependent upside, never the base case), and CPA revenue depends on sportsbook postbacks firing correctly. Partners carry the risk; we carry the responsibility for verifying them — which is precisely what tech development leadership is funded to do.
In 2026 the ad platforms have absorbed targeting, bidding, and optimization — so money and attention go to creative volume, a clean data backbone, and an agent layer that runs the ops. Nothing needs an engineering department; a few pieces need one technical owner.
Acquisition, lifecycle, agentic ops, data backbone, compliance, and B2B GTM land the software floor around $1.8–2.8K/month — the entire "operating department," rented not hired, and fully swappable. The three growth capabilities (audience intelligence, automated GTM, data monetization) are licensed separately — no single point of failure, no supplier grading its own performance.
| The GTM plan commits to… | …and the operating model is why that's achievable |
|---|---|
| A 20,000-profile base in 60 days, 6,000 funded by Month 8 — on a team of two | The agentic stack does the work of an ops + ad-ops + lifecycle + analyst department at a ~$2–3K/mo software floor. The ~$45–60K/mo of payroll we don't carry becomes the media budget that buys those onboards. |
| Two funnels in parallel — B2C consumer and B2B venue/partner | Platform pillars carry both motions on one data backbone — which is why two motions don't require two teams. |
| Activation deferred to Month 3 — Months 1–2 are a deliberate asset build | Revenue arrives in Month 3, so the data backbone must be wired inside the first 60 days with no revenue yet to justify it — precisely why technology leadership is a Day-1 engagement. |
| Illinois → Ohio → North Carolina, sequenced by handle × bar density | WagerWire's licenses across all 35 legal states mean the map is sequenced purely by venue economics. Licensing is a launchpad, not a gate; prediction markets extend the front door to all 50. |
| $107 / $75 / $32 unit economics per activation | Unprovable unless the sportsbook postback fires. Constraint 3 gates all acquisition spend on one validated end-to-end test conversion — tech development's first deliverable. |
| A data-monetization line as the second revenue stream | The clean room sits inside the warehouse we already run. The binding constraint isn't technology, it's consent captured at the QR — a Month-1 design decision that cannot be retrofitted later. |
| Chicago validated, then replicated into 15+ venues and chains | Partners supply warm foot traffic we never bought; the automated-GTM platform runs the venue pipeline so replication doesn't require a BD headcount per market. |
Cumulative revenue against cumulative program cost across three years, on the GTM model's unit economics ($107 / $75 / $32) and the real sports calendar. The first eighteen months are the proving ground; the profit curve is a Year-3 story, when national scale and the platform layer land on a base that's already built. Directional, not a forecast.
| Phase | Registrations | Activations | Revenue | − Program cost | = Contribution |
|---|---|---|---|---|---|
| Months 1–2 · MVP — free adoption | building base | 0 | $0 | build + pilot | $0 |
| Months 3–8 · Launch — activate | 20,000 | 6,000 | $642K | $450K | $192K |
| Months 9–18 · flywheel (~6%/mo) | 73,500 | 22,000 | $2.35M | $1.65M | $704K |
| Months 19–36 · national + platform layer | 966,500 | 290,000 | $31.0M | $21.8M | $9.3M |
| 3-year arc | 1.06M | 318,000 | $34.0M | $23.9M | $10.2M |
We ran an adversarial review of our own plan before writing this. Where the honest answer is "we don't know yet," that is the answer.