Marketing Spend as Perverse Incentive

The spend defends itself

Ryan Cohen cut roughly 800 million dollars of SG&A out of GameStop, 47 percent, "largely by making marketing more efficient, almost turning off marketing."¹ He gives the figure in an interview on buying eBay, and the reasoning carries between the two companies: the brand is already known, "everyone knows GameStop, everybody knows eBay," so the marginal advertisement mostly reaches people who already know the brand and would transact anyway, leaving the incremental return low or negative.¹ Behind the arithmetic sits a principal-agent failure specific to established brands, in which much marketing spend is not buying incremental demand and persists because the people and agencies attached to it are incentivized to protect the budget rather than the profit.

The second half of his claim is about incentives. He reports that "you talk to the marketing people and they'll tell you it's going to tank revenues," but that "most of that marketing spend isn't making money, everyone's trying to protect their jobs, and there's kickbacks. There's all kinds of perverse incentives." In his framing the marketing staff and their agencies are optimized to keep the budget rather than to maximize the firm's profit, which is the classic principal-agent gap between an owner and the people spending the owner's money. He extends the same reading to eBay, which he says "spent 2.5 billion dollars to grow 1 million users," and where he sees roughly 2 billion dollars of cuts available across sales, marketing, and corporate overhead, treated as fast rather than multi-year.¹

Why an owner cuts what a manager protects

Cohen's version of the pattern is inseparable from ownership. His argument is that a salaried marketing organization will defend its own spend, while an owner with concentrated skin in the game reads the same budget as a fast lever on earnings. This is the concrete expression of the broader contrast in owner-operator vs professional management: the cuts are the mechanism by which Cohen quickly raises earnings and pays down the leverage taken on to do a deal, which ties it directly to Capital Allocation Discipline.¹

He generalizes the lean instinct through the Twitter take-private, arguing that deep headcount cuts did not break the service and that "the fewer people you have, the more it's like a startup." He attributes Twitter's commercial damage to an advertiser "conspiracy" rather than to the cuts themselves, which is his own interpretation rather than an established fact and reads as convenient to his cut-deep thesis.¹

The inverse case and the risk

The pattern is sharpened by its mirror image. The same distrust of the agency's incentive can produce the opposite decision, an owner who pulls acquisition spend in-house and feeds it rather than shutting it off, and the reader is directed to the file on the owner-operated growth lever for that version. Both act on one premise, that the spend lever belongs in the owner's own hands. What separates them is whether the advertising is buying demand that would not otherwise arrive. Cohen's whole case rests on the brand already being known, which is what licenses "everyone knows GameStop, everybody knows eBay" and lets him treat the marginal advertisement as waste.¹ Reverse that premise and the identical reasoning argues for spending more rather than less.

The open risk is that "almost turning off marketing" assumes the brand's demand is self-sustaining. For a marketplace, top-of-funnel acquisition may matter more than for a single-name retailer, and cutting it could quietly erode the user base that the durability thesis depends on. The pattern identifies a real source of waste, but it does not by itself distinguish protective spend from the genuine acquisition a brand still needs.