Cutting Your Platform Fee Doesn't Buy You Better Apps. It Buys You More Bad Ones.

Based on the research ofAnnamaria Conti and Juan Santaló, "The Hidden Costs of Fairness in Platform Markets: The Dynamics of Lowering Developer Royalty Rates," Management Science, 2026

For a decade the 30 percent app-store commission has been the most resented number in the digital economy, cast as a tax that starves the little developer to feed the trillion-dollar gatekeeper. So when Apple and Google both moved to halve it for their smallest complementors, the change read as an unambiguous act of fairness: keep more of what you earn, build better things, everybody wins. Annamaria Conti and Juan Santaló, writing in Management Science, went looking for the "build better things" part of that story and could not find it. What they found instead is that lowering the royalty rate drew in a wave of lower-quality, less original apps, worsened the congestion that already buries good software, and quietly raised the price of being seen. The fairness move had a hidden cost, and the platform, not the headline, paid it.

The fairness cut, and what walked through the cheaper door

The natural experiment is unusually clean. On November 18, 2020, Apple announced its App Store Small Business Program, cutting its commission from 30 percent to 15 percent for developers earning up to 1 million dollars in annual proceeds, effective January 1, 2021. Four months later Google matched it, announcing in March 2021 that it would charge 15 percent rather than 30 percent on the first 1 million dollars of revenue every developer earns each year, a change Google said would give 99 percent of its monetizing developers a 50 percent reduction in fees. Two platforms, two near-simultaneous announcements, both aimed squarely at the small end of the developer distribution. That is a difference-in-differences design handed to researchers on a plate, and Conti and Santaló took it.

The intended logic was that a lower take rate raises the return on developer effort, so developers put in more of it and ship better products. The first half held: developers affected by the cut did respond, and they shipped more. The second half inverted. Across both iOS and Android, the authors find a substantial increase in small developers' likelihood of releasing lower-quality and less original apps after the announcements. Crucially, that increase was not matched by any comparable rise in high-quality releases, nor by any greater tendency of small developers to launch the kind of venture-backed apps that signal real ambition. The extra supply the fee cut summoned was concentrated almost entirely at the bottom. Halving the fee did not lift the ceiling on what small developers built; it lowered the floor on who bothered to build at all.

The mechanism is not mysterious once you see it as a selection problem rather than an effort problem. A 30 percent commission is a filter. It quietly discourages the marginal developer whose app was only ever going to earn a little, because after the platform's cut there was not enough left to justify the work. Drop the commission to 15 percent and that marginal project suddenly pencils out. The people most moved by the change are, by construction, the ones who were closest to not shipping in the first place, and the apps closest to not being worth shipping are rarely the good ones. Lower the price of entry and you do not just get more of the same mix. You get a mix weighted toward whatever the old price was keeping out.

The congestion tax nobody put on the invoice

An influx of mediocre apps would be a manageable problem if catalogs were infinite and attention were free. They are not. An app store is a discovery engine with brutally scarce shelf space, and every additional low-quality title makes every other title fractionally harder to find. Conti and Santaló name this directly: the fee cut worsened information congestion, the condition in which the sheer volume of offerings degrades everyone's ability to locate the good ones. The cost of the fairness move did not land on Apple's or Google's margin. It landed on the discoverability of the entire catalog, including the high-quality apps that had nothing to do with the reform.

The paper's most telling evidence is indirect and, for that reason, hard to argue with. If congestion genuinely worsened, then anything that helps an app cut through the noise should have become more valuable after the cut, and that is exactly what the authors observe. The marginal value of experienced-user certifications rose. So did the payoff to app-store optimization, the search-engine-optimization of the mobile world. When getting noticed starts paying better, it is because getting noticed has gotten harder. The fee cut turned visibility itself into a scarce good, and rewarded the developers willing to spend on gaming the ranking rather than on building the product.

A 30 percent commission is a filter. Halve it and the developers most moved to ship are, by construction, the ones the old price was keeping out.

That is the counterintuitive core of the finding. The platform gave money back to small developers in the name of fairness, and the aggregate effect was to make the marketplace noisier and to shift the returns toward promotion over quality. Both giveaways, in other words, financed the same outcome: a catalog that is larger, cheaper to enter, and harder to shop.

Why everyone reached for 15 percent anyway

None of this means the fee cuts were irrational, because the pressure that produced them had almost nothing to do with catalog quality. Apple did not wake up feeling generous. It was being sued, investigated, and publicly flayed over the 30 percent rate, and the Small Business Program landed in the middle of that fight. Ten months after the program took effect, in September 2021, Judge Yvonne Gonzalez Rogers ruled in Epic Games' antitrust case against Apple, siding with Apple on nine of ten counts but issuing an injunction against Apple's anti-steering rules under California's Unfair Competition Law; the Ninth Circuit largely affirmed that outcome in April 2023. The 15 percent tier was as much a political and legal pressure valve as a growth strategy, a way to answer the "you are strangling small developers" charge without touching the 30 percent rate on the large ones who actually generate the revenue.

That framing is what makes the Conti and Santaló result so useful, and so easy to miss. Judged against the political goal, the fee cut worked: it looked fair, it protected the headline commission on the top of the distribution, and it took some heat off. Judged against the ecosystem goal the platforms also claim to care about, a higher-quality catalog and a better user experience, it backfired. The two objectives point in opposite directions, and a reform optimized for the first quietly damaged the second. A platform that cuts fees to win an argument about fairness should not be surprised when the thing it was not optimizing for gets worse.

The generalizable warning is that "lower the fee you charge your suppliers" is never a pure transfer. Your take rate is doing double duty as a screen, and its level is part of how you govern quality whether you intended it that way or not. Cut it and you have loosened a control you may not have realized you were holding.

The curation bill comes due

Watch what the platforms actually spend on after they cut the fee, and you can see them paying the congestion bill in real time. Apple's own transparency reporting shows the scale of the filtering now required to keep the catalog usable: in 2023 the company rejected more than 1.7 million app submissions, including over 248,000 turned away specifically for spam, blatantly copying other apps, or otherwise misleading users, and removed or rejected another 40,000 from developers running bait-and-switch schemes. Those numbers are the downstream cost of an easy front door. When entry gets cheaper, the marginal submission gets worse, and someone has to stand at the gate rejecting the flood, an expense that never appears next to the commission line but is every bit as real.

This is the trade the research forces into the open. The fee is not merely the platform's revenue; it is one lever in a system that also includes review, ranking, certification, and moderation. Pull the fee lever toward "fairer" and you increase the load on every other lever. Cutting the commission without simultaneously tightening curation is not generosity. It is deferring a cost from the income statement to the catalog, where it shows up as noise, as gamed rankings, and as good apps that users never find.

The managerial takeaway is not "keep your fees high." Plenty of good reasons exist to lower a take rate, from competitive pressure to regulatory survival to genuine goodwill, and the political logic behind Apple's and Google's cuts was sound on its own terms. The takeaway is that a fee cut is a quality event, not just a pricing event, and it must be governed as one. If you are going to widen the door, you have to raise the standards for what gets through it, invest in the ranking and trust signals that help quality surface, and budget for the curation load that a cheaper entry point guarantees. The platforms that cut fees and did nothing else did not buy themselves better apps. They bought themselves more bad ones, and the hidden cost of that fairness is still being paid, one buried five-star app at a time.