More Control Over Your Ad Load Is Making You Poorer

Based on the research ofWei, Tian and Jiang, "EXPRESS: Ad Intensity Policies and Content Provision on Revenue-Sharing Content Platforms," Journal of Marketing Research, 2026

Wei, Tian, and Jiang build a model with one platform and two competing content creators and let the platform choose how advertising load gets set: differentiated advertising (DA), where the platform sets a different ad rate for each creator; uniform advertising (UA), where the platform imposes one ad rate on both; or creator-set advertising (CA), where each creator picks their own. The intuitive ranking is CA first, since autonomy should let creators optimize for their own audience. The model says the opposite under common conditions. When creators are symmetric and marginal ad revenue does not depend on quality, UA beats DA on revenue-sharing rate, content quality, ad intensity, and platform profit, and CA is worse than DA on all four, because self-selected advertising blunts the very competition that drives creators to invest in better content. Three live platform decisions show the same logic already playing out at scale.

Twitch already ran this experiment on itself

Twitch used to pay ads through a fixed CPM, a flat rate per thousand ad views set individually by inventory and audience, which is a differentiated arrangement in substance if not in name: what you got depended on who you were. In June 2022, Twitch's own blog announced a shift away from that structure. Partners inside the Ads Incentive Program, and starting that August, any qualifying Affiliate, could set their Ads Manager to run three or more minutes of ads per hour and receive a flat 55 percent revenue share on every ad that ran, a single uniform rate applied the same way regardless of channel size. Twitch said the change represented a 50 to 150 percent ad-pay increase for the vast majority of creators on the platform. The lever Twitch pulled was not "more ads, please." It was "commit to the same ad density as everyone else, and get the same higher share as everyone else." Streamers who instead kept full personal discretion over how many ads to run stayed on the older, lower, less predictable payout. Twitch built the uniform-rule option; the creators who opted into it were paid better for doing so.

This does not prove Twitch's engineers had read a forthcoming JMR paper. It shows the underlying economics holding up under real incentives: a platform-imposed, uniform ad standard, adopted voluntarily, out-earned individual discretion for the people who took it.

YouTube took the dial away and told you why

YouTube's ad-revenue split is itself uniform by design: creators who turn on the Watch Page Monetization Module get 55 percent of net ad revenue on long-form video, a fixed rate that does not vary by creator or negotiate up or down. But the more telling move came in September 2023, when YouTube announced it would stop letting creators choose whether their pre- and post-roll ads on long-form video were skippable or non-skippable. Creators could still decide whether ads ran before or after a video at all, but the format itself would now be set by YouTube's own optimization rather than by creator preference. YouTube's own blog post, reported by Marketing Brew, said the change would barely register for most channels because more than 90 percent of videos already had every ad format turned on, and that an internal experiment found creators running pre-roll and non-skippable ads alongside other formats earned more than creators who left those formats off. Mid-roll placement stayed creator-controlled, manually or automatically, but the freedom to withhold ad formats individually narrowed. YouTube's own data pointed the same direction as the model: the creators exercising the most restraint over their ad load were the ones earning less.

The freedom to run fewer ads than everyone else is not free. Someone is paying for it, and in these platforms, it is usually the creator holding the dial.

TikTok never installed a dial in the first place

TikTok's Creator Rewards Program, launched out of beta in March 2024, gives creators no lever over ad load at all. TikTok's own newsroom post describes an "optimized rewards formula" built from four factors it scores algorithmically: originality, play duration, search value, and audience engagement, plus an automatic adjustment for "the account's ad value determined by their community's ad watchtime." Eligibility requires 10,000 followers and 100,000 views in the trailing 30 days. Nowhere in the program does a creator choose how much advertising runs against their content; TikTok's system decides that centrally and folds it into a single payout formula applied the same way across the eligible pool. It is closer to a uniform, platform-set policy than to anything resembling creator-set advertising, and it launched at a moment when TikTok said its community was already spending half its time watching videos longer than a minute, the exact inventory ad-intensity policy is built to monetize.

Why the theory is more careful than the headline

The paper is not making a blanket case against creator autonomy. The advantage of UA over DA and CA is explicitly conditional: it holds when creators are symmetric and marginal ad revenue does not depend on content quality. Loosen either assumption and the ranking can flip. When creators are unevenly matched, or when higher-quality content pulls in disproportionately more ad revenue per impression, a differentiated policy can out-earn a uniform one, because forcing an identical ad rate onto unequal creators wastes the extra value a stronger creator's audience could generate. Creator preferences split the same way: when creators are close substitutes for one another, UA tends to benefit them because intensified competition still lifts the whole market, but when creators are weak substitutes, that same competitive intensification can hurt the weaker or more differentiated creator, who would rather have kept a rate suited to their own audience. The honest reading is not "uniform rules always win." It is that individual choice, framed as fairness, is not a proxy for what maximizes payout for anyone; the right policy depends on how similar the creators are and where the ad dollars come from, and platforms need to check those conditions before copying a rule that worked somewhere else.

The takeaway for anyone designing a revenue-sharing rule

Every ad-intensity choice a platform makes is really a choice about who absorbs risk and who captures the upside of quality competition. A uniform rule concentrates that competition and, under the right conditions, raises the payout for everyone inside it, which is what Twitch's flat 55 percent and YouTube's fixed 55 percent watch-page split are already doing in practice. A creator-set rule feels fairer in the moment, but it lets each creator quietly under-invest relative to what a shared standard would have pushed them toward, which is the trap TikTok's algorithmic, non-negotiable payout formula sidesteps by never offering the choice at all. None of this licenses removing every lever creators hold; it means testing, before removing one, whether your creators are similar enough and your ad economics flat enough for uniformity to be the generous choice rather than the restrictive one.