Alignment Architecture: the next thing to build in agriculture

iSelect forward thesis. Grew out of the Daws/Alpert debate.

← Concepts Index · Home · The PPA Pattern · Incumbent Capture

The thesis in one line

In agriculture, the science already works. What's missing is the alignment architecture — the contracts, offtakes, and financing instruments that let a proven technology reach the market at a value-added price instead of leaking its value to the incumbent. Building that architecture, not funding another capability, is the highest-return move in the sector.

The problem, stated plainly

Science creates the possibility of value. Alignment determines who captures it. In most industries those two happen close together. Agriculture has one of the longest, most fragmented value chains in the economy — grower, elevator, crusher, processor, brand, retailer, payer — so a technology can be exceptional and still fail, because value capture depends on coordinating dozens of independent participants who are each priced on commodity inputs, not outcomes.

The result of the last decade: the science got funded and de-risked, then the value got captured by strategics buying cheap (Blue River → Deere, Climate Corp → Bayer) or never got captured at all (Benson Hill had the high-oleic science; what it lacked was a durable offtake, later-stage capital fluent in the risk, and a market that priced nutrition quality instead of input cost). None of those were technology failures. Every one was an alignment failure.

What "alignment architecture" actually is

Not a metaphor. A concrete stack of instruments:

• Value-added offtake — a long-term contract that pays a grower or processor a premium for a measured attribute (nutrient density, oil profile, carbon, protein quality), the way a solar PPA pays for delivered kilowatt-hours. This is the missing primitive. There is no PPA for nutrient density today.
• Outcome-based demand — a downstream counterparty (self-insured employer, payer, CPG, health system) contracted to buy the outcome, so the premium has somewhere to land.
• Purpose-built capital — first-loss / blended tranches and later-stage vehicles that understand food-system risk, so a company doesn't die in the Series C valley the moment it needs to scale production.
• Coordination layer — the connective tissue (data, standards, traceability) that lets the chain verify and settle the attribute being paid for. (This is the strongest version of Alpert's "architecture" — and it's a capability someone has to fund and build, not just screen for.)

The pattern is not new — every stalled category waited for a financial realignment, not a better technology

In each case a financial or business-model instrument, not a new invention, moved the category. Agriculture is the same story with the instrument still unbuilt.

Why the seam is defensible

The instrument goes unbuilt because it lives in a seam almost no one occupies. People who can structure capital have near-zero tacit knowledge in agriculture. People with the tacit knowledge — growers, processors, agronomists — have no access to capital-markets tools. The alignment architecture requires both at once. That intersection is thin, hard to fake, and compounds with reputation and relationships — which is exactly what makes it a moat rather than a feature. iSelect's edge is sitting in that seam: capital-markets structuring fused with real ag domain depth and an operator network.

What we underwrite differently

1. Evaluate and build alignment architecture with the same rigor the market applies to technical risk — but don't stop at diligence; construct the missing offtake/financing pieces as part of the investment.
2. Optionality over ownership. Diversify early to see live signal; hold participation rights; concentrate only when the alignment path is proven (the Optionality structure).
3. Price to the right denominator. The 10x was never in the ~$0.5T farm gate at commodity margins. It's in a value-added, nutrient-density food system taking share from a $5.7T healthcare spend. Alignment architecture is what lets that value flow back to the companies that create it. See The Prize.

The honest part (so this is auditable, not a pitch)

• DPI is the scoreboard. Spillover and "the science works" are not a substitute for returns — they're the reason the next return is cheap to finance, because the last cohort's dead equity already paid for the de-risking. We are not asking to be graded on vibes.
• The instrument doesn't fully exist yet. That's the ask and the roadmap, not a finished claim. Proof-of-life is what we build toward: a first value-added offtake signed, a first payer/employer underwriting an outcome, a first later-stage tranche structured against it.
• GLP-1s are the demand shock, not the proof that food wins. A drug captured the first metabolic-health dollar. The food opportunity is the durable pull the drugs create — reformulation, nutrient-dense demand, the populations drugs don't reach. If pharma keeps capturing the health dollar and food stays an input, we're wrong; the wedge is real where drugs are weak.

Falsifiable milestones (what would prove — or kill — this)

• 12 mo: at least one value-added offtake structured with a named downstream counterparty; if none is signable, the "no PPA exists" gap is a market preference, not a missing instrument — and the thesis weakens.
• 24 mo: a later-stage capital source underwriting food-system risk on our terms (proof the Series C valley is bridgeable).
• 36 mo: measurable value flowing to a portfolio company from a health/outcome counterparty — the first data point that food is capturing healthcare share, not just asserting it.

One line to leave them with

Science creates the possibility. Alignment architecture decides whether the market — or the strategic — captures the value. That architecture is the highest-return thing left to build in agriculture, and it lives in the seam between capital markets and the field, where almost no one sits.