Independent estimates converge near $3 billion for ex-China dramas

The figure or the record

Media Partners Asia (MPA), a regional research firm, published a report in August 2026 titled ReelShort / Crazy Maple Studio: Inside the US$1B Micro-Drama Machine, which states an 'ex-China micro-drama market sizing: US$2.7 bil. in 2025 and US$3.6 bil. in 2026', projected to reach 'US$9.5 bil. by 2031'. The page names its own method: the estimate is 'anchored to MPA's proprietary micro-drama market model, structured discussions with CMS [Crazy Maple Studio] management, consumer panels operated by MPA and its subsidiary ampd Analytics, and regulatory filings'.

What the documents establish

MPA states plainly that this is 'an independent MPA assessment', built on a named model, management interviews and its own consumer panel — a modelled estimate, not a measured transaction total. The figure sits close to a separate estimate: Owl & Co's Short Report put ex-China short-drama app revenue at about US$3 billion for 2025, against MPA's US$2.7 billion for the same year — a few hundred million dollars apart despite different firms, panels and models. MPA's own reports page also lists a separate, broader 'Micro-Drama Economy 2025' report, confirming the firm treats the sector as large enough to support more than one paid study.

Scope and caveats

The headline figures are ex-China only, and the report page does not state whether they cover in-app purchases alone, or also web payments and advertising; the model runs 'through 2028, with scenario analysis to 2030', so the 2031 figure is a projection, not a base-year measurement. The report is built substantially around one company, Crazy Maple Studio (ReelShort's operator); MPA's page does not disclose what share of the ex-China total that one company represents. MPA states its own limits directly: 'this report is an independent MPA assessment. It does not constitute investment advice.'

What to watch

Editorially, the useful check is whether MPA's 2026 figure (US$3.6 billion) and Sensor Tower's Q1 2026 quarterly IAP run rate move in the same direction once a full year of 2026 data exists, since the two firms measure different things — a modelled annual total against store-observed quarterly purchases.

• Does a quoted ex-China micro-drama total include advertising and web payments, or in-app purchases only?
• How much of a market-wide estimate rests on one company's data, and is that share disclosed?
• Is a multi-year projection presented with the same confidence as the base-year figure it extends from?

Two independent estimates landing within a few hundred million dollars of each other is a reasonable cross-check, not proof; both are models built substantially on the same handful of visible companies, so agreement between them may reflect a shared blind spot as much as an accurate market.