← Back to the interactive chart · 中文
About & Methodology
Where Money Goes is an interactive map of the world economy. You start from roughly $118 trillion of global GDP and click your way down — through household spending, entertainment, digital purchases and app stores — until you reach a single mobile game. Every level shows how big a slice is, what share of its parent it represents, and where the number comes from.
Data year: 2025 · Last updated: October 10, 2026
The main path, step by step
| Level | Size | Share of the level above |
|---|---|---|
| Global GDP | $118.2T | — |
| Household consumption | $66.9T | 56.6% of GDP |
| Recreation & culture | $4.7T | 7.0% of household consumption |
| Direct digital spending | $430B (est. $380–480B) | 9.1% of recreation & culture |
| In-app purchases (incl. subscriptions) | $167B | 38.8% of direct digital spending |
| Mobile game in-app purchases | $81.8B | 49.0% of in-app purchases |
| Tencent (store-reported mobile revenue) | ≈ $10.5B | ≈ 12.8% of mobile game IAP |
| Honor of Kings | ≈ $2.4B | ≈ 23% of Tencent's figure |
Put differently: one hit mobile game is about 0.002% of world GDP — yet it is still a multi-billion-dollar business. That contrast between scale and detail is what the chart is built to show.
How to read the chart
- Main path (blue/green/purple tones). Every slice is part of its parent, so the children of a node always add up to the parent. You can multiply the shares along a path to get a share of GDP.
- Side views (orange, dashed outlines). Some famous numbers — global digital advertising, e-commerce GMV, platform commissions and app-store fees — are not slices of the main tree. They are shown next to the node they relate to and can be explored on their own, but they must not be added to the main path.
- “Other / unspecified” (gray). When we list only the largest items under a node, the remainder is calculated as the parent minus the listed children. This keeps areas honest instead of inflating the named items.
- “Est.” badges. Many sub-items are order-of-magnitude estimates. They are labeled so you always know which figures are official statistics and which are reasoned approximations.
Accounting rules we follow
GDP by expenditure
The top level splits GDP the way national accountants do: household consumption + government consumption + gross capital formation + net exports. Globally, net exports should be zero; it is slightly positive in published data because of statistical discrepancies between countries.
Household consumption categories
Household spending is grouped using the UN Classification of Individual Consumption According to Purpose (COICOP): housing, food, transport, health, recreation and so on. Global shares are estimated by combining national accounts from the largest economies, so they should be read as approximate.
Imputed rent
About $7.5T of “housing” is rent that homeowners notionally pay themselves. National accounts include it so that GDP does not change just because more people own rather than rent. It is the single largest item in household consumption, which often surprises people.
Why advertising is not part of consumer spending
Advertising is paid for by businesses. In national accounts it is an intermediate input: its cost is already embedded in the prices of the goods and services households buy. Adding the ~$715B of digital advertising on top of consumption would count the same dollars twice, so it appears only as a side view.
Why e-commerce GMV is a side view
Gross merchandise value (~$6.4T) measures what passes through online checkouts. It includes business purchases and returns, and it overlaps with many household categories (food, clothing, electronics). It describes a channel, not a separate slice of spending. Platform take — commissions and fees — is in turn part of GMV.
De-duplicating digital spending
Subscriptions can be bought inside an app or on the web. We count in-app purchases once (from app-store data) and list outside-app subscriptions separately with in-app revenue removed, so Netflix or Spotify revenue is not double counted.
Sources
- Global GDP: IMF World Economic Outlook, October 2025 (nominal US dollars).
- Household consumption share: World Bank World Development Indicators, household final consumption expenditure.
- App and mobile game spending: Sensor Tower 2025 reports (App Store + Google Play).
- Military spending: SIPRI Military Expenditure Database.
- Companies and titles: annual reports, earnings releases and public app-store rankings, converted to US dollars and rounded.
- Digital advertising and e-commerce: consensus of published 2025 industry estimates.
Limitations
- Mobile game figures cover iOS and Google Play only. Third-party Android stores in China are excluded, so Chinese publishers such as Tencent and NetEase are understated.
- Company figures sometimes mix consumer and business revenue (for example, automaker revenue includes fleet sales). Where that happens it is noted on the node.
- Different sources use different fiscal years and exchange rates. Numbers are rounded to keep the picture readable; treat them as orders of magnitude, not audited statistics.
Spotted a number that looks wrong, or have a better source? Email [email protected]. Corrections are welcome and will be credited.
关于与方法说明
Where Money Goes（全球经济钻取）是一张可交互的世界经济结构图。从约 118 万亿美元 的全球 GDP 出发，逐层点击——居民消费、娱乐与文化、数字直接付费、App 内购——一直钻到某一款手游。每一层都显示规模、占上一层的比例和数据来源。
怎么读这张图
- 主路径：每个扇区都是上一层的一部分，子项相加等于父项，沿路径把占比相乘就能得到占 GDP 的比例。
- 旁支（橙色虚线）：数字广告、电商 GMV、平台抽成、应用商店抽成等口径不属于主路径，可以单独钻取，但不能与主路径相加。
- 「其他 / 未细分」（灰色）：只列出最大的若干项时，剩余部分 = 父项 − 已列子项，避免夸大具名项目。
- 「估算」标记：大量细项是量级估算，均已标注。
核算口径
- 支出法 GDP：居民消费 + 政府消费 + 资本形成 + 净出口。全球净出口理论上为 0，公开数据因统计差异略为正。
- 消费分类：采用联合国 COICOP 分类，全球份额由主要经济体的国民账户汇总估算。
- 推算租金：约 7.5 万亿美元的住房支出是自有住房「自己付给自己的房租」，这是居民消费中最大的单项。
- 为什么广告不算消费：广告费由企业支付，属于中间投入，已经体现在商品价格里，加到消费上会重复计算。
- 为什么电商 GMV 是旁支：GMV 是成交渠道而不是独立的支出类别，包含企业采购和退货，并与食品、服装、电子产品等类别大量重叠。
- 数字付费去重：App 内购按应用商店数据只算一次，App 外订阅已扣除 App 内部分。
数据来源
IMF《世界经济展望》2025 年 10 月；世界银行 WDI；Sensor Tower 2025；SIPRI；各公司年报与财报；行业机构 2025 年数字广告与电商估算。
局限
手游数据只含 iOS 与 Google Play，不含中国安卓渠道，腾讯、网易等被低估；部分公司数据混有企业端收入；不同来源的财年与汇率不一致。所有数字均为量级参考，不是审计过的统计数据。
发现数据有误或有更好的来源，欢迎发邮件到 [email protected]。