Act A — The Margin in the Middle
Saskatchewan grows more lentils than almost anywhere on earth.
The province exports roughly 90 percent of what it produces. Most of it moves through commodity brokers to South Asian and Middle Eastern buyers. A small but growing portion goes to US food manufacturers. Almost none goes directly to Mexico, despite Mexico's enormous and growing legume-based food industry — canned beans, instant soups, health-bar protein fillings, and traditional salsas that have used lentils and chickpeas for generations.
The reason is simple: the connection doesn't exist.
Canadian pulse producers sell into commodity channels because that's the infrastructure they have. Mexican food processors buy through US brokers because that's the only channel they know. The US broker earns the margin, sets the specification, and owns the relationship. Both the Canadian producer and the Mexican processor are price-takers in a transaction that could, in principle, be direct.
The following is a fictional account of how a MarketForge-powered bilateral trade platform changes that equation for one co-operative and one processor.
Act B — The Story
Darren manages a pulse producer co-operative outside Moose Jaw with eleven member farms. Three consecutive strong harvests have left the co-op with more red lentils than their existing broker channels can move at price. Darren has heard that Mexican demand for lentils is growing — he read it in a Pulse Canada market brief — but has no idea how to reach a Mexican buyer directly. He doesn't know what COFEPRIS is. He doesn't know whether CUSMA makes the tariff workable. He registers the co-op on the MarketForge bilateral trade platform after it's mentioned in a Saskatchewan Pulse Growers webinar.
The onboarding process asks about variety, CGC grade, protein content, moisture, harvest year, available volume, and whether the co-op has an existing customs broker. Darren answers what he can, leaves the customs broker field blank, and notes that the co-op has ISO 22000 food safety certification.
Sofía is the procurement director at a mid-sized food processing company in Guadalajara that produces canned legume products for Mexican grocery chains and a growing natural-food export line. She buys red lentils through a US commodity broker in Dallas at prices that have been climbing for eighteen months. She pays a brokerage margin she finds increasingly difficult to justify to her CFO, and she has no ability to specify the protein content tier she actually needs — the broker ships what it ships.
Her company registered on the platform at the suggestion of a CANACINTRA trade association contact.
The platform's matching algorithm surfaces Darren's co-op against Sofía's demand profile. The protein content range, moisture specification, and CGC grade align with her requirements. The volume is workable for a trial shipment. The platform generates a match notification for both.
Along with the notification, the platform generates a Generative Match Story — a structured description of how this specific trade could be arranged, drawn from both parties' profiles and KnowledgeSlot's curated knowledge of the CUSMA pulse trade corridor.
The scenario describes the CUSMA certificate-of-origin documentation Darren's co-op will need to generate to qualify for Mexican zero-tariff treatment. It explains the COFEPRIS import notification process Sofía's company will need to initiate — a relatively simple process for food-grade pulses, but one her procurement team has never completed. It identifies two established customs brokers with offices at Laredo who handle Saskatchewan-origin agricultural product regularly. It describes the logistics sequence: Saskatchewan elevator to rail to Laredo, customs clearance, Mexican trucking to Guadalajara.
Sofía's first reaction is surprise. She didn't know the tariff was zero. She assumed Canadian lentils would be subject to import duties that made them uncompetitive with US-sourced product. They aren't.
Darren's first reaction is that the customs broker list is the most useful information he's received in three years of considering direct export. He calls the first broker on the list that afternoon.
They execute a trial shipment agreement for one container — 22 tonnes of red lentils, CGC No. 1 Small, protein 26.5% minimum — eight weeks later. Sofía specifies the protein floor. Darren's co-op meets it. The trial shipment arrives in Guadalajara within spec.
Sofía's CFO approves a standing annual supply agreement for the following season.
Act C — Why This Market Stays Broken Without Infrastructure
The structural barrier in this scenario is not price, distance, or quality. It is the absence of any shared discovery channel between Canadian pulse producers and Mexican food processors.
Both parties want the outcome. Both are commercially ready to execute. The gap is a specific set of obstacles — CUSMA documentation, COFEPRIS process, customs broker identity — that neither party can resolve efficiently without someone who knows both sides of the corridor.
US brokers fill this gap today. They fill it expensively, and they fill it in a way that gives neither party a direct relationship with the other. The broker is the infrastructure, and the broker charges accordingly.
What thin market infrastructure does here is replace the broker's information advantage with a shared platform that both parties can access. The CUSMA knowledge, the COFEPRIS process, the logistics network — all of it made available to the parties directly, at the moment they need it, without the margin extraction that a traditional intermediary requires.
The characters and companies in this story are fictional. Trade data and regulatory frameworks described — CGC grading standards, COFEPRIS import procedures, CUSMA certificate-of-origin requirements — are real. DeeperPoint is a real project building the infrastructure this story describes.