The Brown Family Controls the Vote. Canada Controls the Shelf.

On a cap table, Brown-Forman looks like the finished version of founder control.

The Brown family owns more than half of the company's voting stock. Its voting position can determine every director election and the approval or rejection of a merger, change of control, or other major shareholder transaction. The structure is perpetual. There is no sunset waiting to turn the family into ordinary public shareholders after everyone has behaved responsibly for ten years.^BF10K

Then Canadian provinces removed Jack Daniel's from their shelves.^BF10KLCBOBAN

There was no proxy fight. Several provincial governments told their liquor authorities to stop buying or selling American alcohol. Those directives did not alter Brown-Forman's voting structure. They made its bottles unavailable.

That is the useful governance paradox. Control of a corporation governs decisions the corporation is allowed to make. It does not govern the customer, the distributor, the platform, or the state body standing between the product and the buyer. In some businesses, the most consequential vote is an order that never arrives.

Brown-Forman's share structure derives its durability from what is absent. Its Class A common stock carries full voting power, including the election of directors. Class B generally does not vote. Shareholders approved the two-class arrangement in 1959, and the company's charter and bylaws contain no provision that eventually converts the non-voting stock into voting stock.^BF10K

This is dual-class stock without the usual countdown. Public investors can own the economics through either listed class, while voting control remains concentrated in the Class A holders. Brown-Forman's 2026 proxy reported that two Brown-family-controlled entities owned 119,809,784 Class A shares as of June 9, 2026, equal to 71.1% of the voting class.^BFPROXY

The company therefore qualifies as a controlled company under New York Stock Exchange rules. It can use exemptions from requirements that otherwise call for a majority-independent board and fully independent compensation and nominating committees. Brown-Forman voluntarily maintains a majority of independent directors and an independent compensation committee, while its nominating committee is not fully independent.^BF10KBFPROXY The family has the votes and has elected to add some supervision anyway, which is governance's version of owning the building and agreeing to a very polite tenants' committee.

The control itself is broad. Brown-Forman says the family can determine shareholder votes, elect all directors, and approve or reject a merger or change of control. It also says the structure may make the company a less attractive takeover target.^BF10K That power matters in a takeover or board challenge. Ontario's purchase decisions sit beyond it.

Brown-Forman describes three routes to the consumer outside the United States: owned distribution, third-party partners, and government-controlled markets. It owns distribution companies in several countries and uses contracted distributors in others. Its Canadian description takes one sentence: "In Canada, we sell our products to provincial governments."^BF10K

That sentence carries more governance information than it first appears to. Federal law generally prohibits importing intoxicating liquor into a province unless the liquor has been purchased by or for, and consigned to, the provincial government or an authorized provincial liquor authority, subject to specified exceptions.^IILA For imports governed by that rule, the province occupies the route itself.

Ontario shows the mechanism clearly. At the end of its 2025 fiscal year, spirits could be sold at LCBO stores and LCBO Convenience Outlets, and the LCBO was the exclusive wholesaler serving licensed grocery and convenience stores.^LCBOAR Retail access later expanded, yet the gate remained concentrated. Beginning April 1, 2026, the LCBO became exclusive wholesaler for all Ontario retailers, bars, and restaurants selling alcohol. Authorized distributors can fulfil orders, but they do so on the LCBO's behalf.^{LCBOWHOLESALE} Ontario added thousands of storefronts and kept one front door.

That arrangement can be efficient. A producer gains one large buyer, standardized logistics, and access to an entire province. The same concentration also turns one procurement decision into a market-wide event. A conventional retailer can delist a bottle from its own stores. A provincial authority that is importer, wholesaler, and retailer can reach stores it does not operate.

Founders tend to call this distribution when it works and platform risk after it stops. The contract did not change categories. Their attention did.

On March 4, 2025, Ontario directed the LCBO to restrict all sales and imports of American beverage alcohol in response to US tariffs on Canadian goods. The LCBO ceased purchasing American products. Customers could no longer buy them through its website or app. Grocery stores, convenience stores, bars, restaurants, and other wholesale customers could no longer order them through the LCBO. American products also disappeared from LCBO stores and convenience outlets.^LCBOBAN

The scale belonged to the channel rather than any one producer. The LCBO said it was importer of record for US alcohol products with annual sales of up to C$965 million and more than 3,600 listings from 35 states. Purchases would resume only when the provincial government directed the LCBO to return to normal business.^LCBOBAN The memo had better distribution than the whiskey.

Ontario was one province in a wider response. Manitoba directed Manitoba Liquor and Lotteries to pull American products from Liquor Mart shelves and stop ordering them effective February 4, 2025.^MB Quebec instructed the Société des alcools du Québec to remove American alcohol from stores and stop supplying grocery stores, agencies, bars, and restaurants.^QC A limited Quebec sale in February 2026 covered inventory at risk of deteriorating. The SAQ said it would donate an amount equal to the cost of eligible products sold, up to C$8.6 million, while the broader prohibition on other sales and new American orders remained in force.^SAQ26

Brown-Forman's filing records the commercial consequence in plainer language. Several provinces removed American beverage alcohol, including Jack Daniel's, in March 2025. By the fiscal year ended April 30, 2026, the company was still attributing Canadian volume declines in American whiskey and Jack Daniel's ready-to-drink products to their continued absence from retail shelves in most provinces.^BF10K

For the nine months ended January 31, 2026, Canadian net sales fell 60% on a reported basis and 59% organically from the prior-year period.^BFQ3 The shelf policy did not necessarily cause every point of the decline. Brown-Forman's own explanation establishes the missing shelves as a driver of lower Canadian volumes. The voting structure remained intact while a material share of Canadian revenue vanished.

Calling a provincial liquor authority a super-voting shareholder is a metaphor. The province does not own Brown-Forman stock, appoint Brown-Forman directors, or owe duties to its investors. That is precisely why its power is easy to omit from a governance diagram.

Put the two systems beside each other. Brown-family-controlled entities held 71.1% of the voting class, and Brown-Forman describes the family's power over directors and major shareholder transactions.^BFPROXYBF10K Ontario's government directed the LCBO's purchasing and sales restrictions through a system where the LCBO handled import, wholesale, and retail access.^{LCBOBANLCBOARIILA}

The Brown family has legal control of the corporation. During the restrictions, Ontario exercised practical control over the principal routes carrying imported spirits to Ontario buyers. Each form of control operated within a different perimeter.

This is why customer concentration belongs in a control analysis rather than living only in the risk-factor section. Credit concentration asks whether the customer will pay. Revenue concentration asks how much disappears if the customer leaves. Control analysis asks whether the customer can change the rules for reaching everyone downstream.

The last question matters for app stores, cloud marketplaces, payment processors, hospital purchasing groups, defence procurement, automotive distribution, franchise systems, and any industry where a regulated intermediary sits between supplier and end user. The intermediary may buy little for itself. It can still determine who else is permitted to buy, on which terms, and through which channel.

There may be no board seat, information right, or protective provision to diligence. The leverage can sit in legislation, a licence, a procurement policy, an API permission, or the commercial fact that replacing the channel would take two years. Nobody calls it governance because the customer declined to wear a suit.

The Brown-Forman case is unusually visible because both sides publish their authority. The company describes family voting control in an SEC filing. Provincial governments announced the shelf decisions in press releases. A startup may discover comparable leverage through a policy notice from an account manager.

The underlying pattern is common:

• A marketplace supplies most new customers and can suspend the listing.
• A payment processor can stop settlement while conducting a review.
• A channel partner owns the customer relationship and can favour another product.
• A government buyer can change an approved-vendor list after an election or budget decision.
• A licence holder can determine whether the product enters a country even when consumer demand remains intact.
• A large enterprise customer can require integrations, pricing, insurance, or compliance work that effectively redirects the supplier's operating plan.

None of these stakeholders needs equity to shape the company. Equity is one method of obtaining leverage. Dependency is another.

This distinction should change how founders evaluate control. A voting agreement can protect your board seats while a platform policy makes the company uneconomic. A founder can preserve super-votes through five financings and still build a business whose only viable customer-acquisition channel belongs to someone with a support portal. The cap table is pristine. The login has been disabled.

Concentrated channels can be rational. Brown-Forman reaches Canadian consumers because the provincial systems offer enormous coverage, and a startup may accept a dominant platform or anchor customer to gain speed and distribution. Ask what the concentration purchased and what right the other party gained in return.

A board and shareholder control map records who initiates, approves, and blocks corporate actions. Extend the same method to every route that carries material revenue.

For each important customer, distributor, marketplace, licence, and procurement body, record:

Do this before negotiating another point of voting control. Then stress-test the business against the loss of the largest gate for one quarter and one year. Include the working-capital effect of inventory already produced for the channel, the cost of discounts needed to redirect it, and the time required to obtain a replacement licence or integration.

Keep legal authority and commercial leverage in separate columns. Record any contractual or public-law remedy alongside the time and cash required to pursue it. A company may hold a claim against a distributor and still run out of cash before the dispute restores a viable route to revenue.

Finally, put an owner beside each workaround. "Diversify channels" is a board-deck aspiration. "Secure a second processor before renewal, with 20% of volume live by October" is a control response. If the alternative exists only after the primary gate closes, it is an incident plan being introduced to itself.

Brown-Forman did not misunderstand its own governance. Its filings describe both systems accurately: the family controls the voting stock, and the company sells to provincial governments in Canada. The mistake available to everyone else is reading only the first sentence and calling the company controlled.

Read the revenue path with the same suspicion you bring to a voting agreement. Find every party that can stop the product before it reaches the buyer, even when that party owns no shares and appears nowhere on the board chart. More than 50% of the vote is control of the corporation. Zero shelf space is control of the result.