The Order Book for
Human Innovation

The Graveyard & The Winner

This category has two decades of history, and the record is unambiguous. Every venture that tried to be a listing venue for intellectual property — post assets, wait for buyers — failed, regardless of the technology attached to it. The one venture that reached institutional scale inverted the model entirely: it started from buyer demand and used the marketplace to fulfill it.

The transaction layer is graveyard territory. The order-driven services layer produced the only $B-scale winner.

Seven failure modes — and what each one teaches

1. One-sided liquidity. Sellers always show up; buyers never do voluntarily (the IPwe post-mortem in one line).
2. Litigation is the real marketplace. IPXI closed because licensees preferred to be sued — efficient infringement outcompeted the exchange.
3. Valuation opacity. Ocean Tomo 2009: unrealistic reserve prices, 6 of 81 lots sold.
4. Lemons / adverse selection. Good IP is monetized privately; public listings skew junk-heavy. Only ~21% of brokered packages ever sell; the entire brokered patent market is ~$158M/yr (Richardson Oliver, 2025).
5. Transaction complexity. A patent is not a product — the reason yet2 pivoted from web marketplace to services.
6. Wrong framing. Selling patents (exclusion rights) is not the same as selling solutions to problems.
7. Tech theater ≠ liquidity. IPwe's blockchain and NFTs did not create one buyer.

IEM's architecture is a point-by-point answer to this list: demand first, transparent valuation, verified supply, standardized transactions, solution framing, and technology applied only where it removes cost — never as theater.

What Changed

The failure modes above were structural in 2010–2020. Four shifts since then remove the cost structure that killed every prior attempt — and they arrived together, within roughly the last three years.

The IEM Model

IEM is the first AI-agent-run innovation exchange: a global order book where companies post innovation needs and innovators sell via fixed price, auction, or license slots — with instant AI valuation, agentic due diligence, and escrowed settlement built in. Two pillars, one trust layer.

Post a need. Our agents find, verify, price, and close.

Total published friction is ~10–15% versus 25–40% in the legacy brokered market — and every listing is machine-readable (MCP/API-first), making IEM the first innovation market built for buyer AI agents to shop in.

Competitive Landscape

The full field, from the dead exchanges to the newly funded AI workflow tools. Read the last column: capital is flowing into tools that stop short of running a marketplace.

The structural insight: the transaction layer is graveyard territory (IPXI, IPwe, Ocean Tomo); the order-driven services layer produced the only $B-scale winner (Scientist.com); and 2024–26 capital is flowing into AI workflow tools (Tradespace, Patlytics) that stop short of running a marketplace. Nobody combines AI-agent execution with a two-pillar transactional marketplace. That is IEM's gap.

Ten game-changer differentiators

COMPETITIVE CHECK — no incumbent does ≥3 of these together.

Market

The addressable value flow is enormous and growing; the open, transactable share of it is vanishingly small. The spread between the two is the opportunity.

>$1T of value flow vs a $158M open market — a >6,000× liquidity gap. That gap is the story.

Business Model & Unit Economics

Listing is free everywhere — supply feeds the data flywheel. IEM monetizes success, verification, enterprise rails, settlement, and data. Five revenue streams:

• Success fees — 8–18% by venue. Direct sales 8–15%; auctions ~10% seller + 5% buyer premium; license slots 10–15% of upfront plus 5–10% of running royalties; Innovation Orders tiered 18/14/10% by deal size (HeroX-proven, pay-on-success), plus 5–10% margin on commissioned development.
• Verification tiers. Paid IEM Score deep verification, Tech Passport issuance, and agentic due diligence packages layered above the free baseline.
• Enterprise private order books. $50–250K/yr SaaS for a company's confidential demand channel on IEM rails — the model Scientist.com proved with 12 private marketplaces plus the NIH.
• Escrow spread. Settlement margin against the 0.89–3.25% industry benchmark (Escrow.com).
• Data & insights subscriptions. Buyer-side subscriptions at ~$4,200/yr ARPA — pricing signals, royalty comps, and demand analytics generated by the exchange itself.

Unit economics (business plan)

The Raise & Roadmap

Staged capital against staged proof. The pre-seed exists to validate one thing — that an order book for innovation clears — before venture-scale capital is committed.

Go / No-Go gates

Five-year financial trajectory

Free cash flow runs from −$3.75M in Year 1 to +$19.71M in Year 5, with break-even in Year 4 and a maximum cumulative deficit of $10.74M — inside the staged raise.

Structure & strategy

• Singapore HoldCo. Singapore Pte Ltd holding company with operating subsidiaries — neutral, treaty-rich, investor-familiar. Pilots anywhere; no single home market.
• Vertical beachhead. One industry first — mining / energy / agri-tech families — per the Scientist.com lesson: it took ~10 years and vertical focus to reach $B scale. A vertical beachhead beats a horizontal launch.
• Lead with problems, not patents. Orders open the market; the sell-side follows demand. Auction liquidity is never promised before buyer density exists.

Exit Comps

Acquirers and growth investors are already circling this exact space — the comp set spans validation, caution, and fresh capital:

The pattern: capital rewards transaction infrastructure with demand-side gravity and punishes networks without rails. IEM is built on the winning side of that line from day one.