The Maze: TikTok Shop lets people buy products inside the video app, through recorded videos, live broadcasts or product pages. In Germany, France, Italy and Spain, recorded videos generated a majority of the value of goods sold in 2025. That is a useful distinction for brands deciding how to sell: a content-led marketplace does not automatically require a live presenter. The strongest format can be something shoppers watch whenever it suits them.
Video is the main selling route in all four countries. The GMV-weighted aggregate puts video at 64%, live broadcasts at 17% and product cards at 19%. Gross merchandise value, or GMV, means the value of goods sold, rather than the marketplace's revenue or sellers' profit. Weighting gives larger markets more influence, so 64% is not a simple average of four national percentages. For a brand, the useful starting point is the scale of video-led shopping. It does not follow that 64% of its own budget should go into videos: spending, audience and production costs are absent from this comparison.
Germany needs a different format mix from its neighbours. Video contributes 69.3% of merchandise value there, compared with 62% in France, 61% in Italy and 59% in Spain. The German lead over Spain is 10.3 percentage points. Live contributes just 12.9% in Germany, versus 20% in both France and Italy and 19% in Spain. A regional plan that gives live the same role everywhere would overlook this difference. A seller could begin with more recorded demonstrations in Germany while testing live formats elsewhere. That is an operating hypothesis to test, rather than a return guaranteed by these market shares.
A product page still has a commercial job. Product cards are listings with a picture, price and purchase button. Their shares exceed live in Germany, at 17.8% versus 12.9%, and Spain, at 22% versus 19%. Product discovery and product evaluation need not happen in the same place. A demonstration can attract interest while a clear listing explains exactly what the buyer gets. The evidence does not track that customer journey, but it warns against treating listings as irrelevant simply because the surrounding app is built around entertainment. Sellers still need an understandable offer when attention turns into purchase intent.
Recorded content and live events demand different work. A recorded demonstration can remain available after publication; a live sale asks the host and audience to be present together. That gives a seller two different production problems: building reusable product explanations, and staffing scheduled events. The ECDB analysis links this distinction to the observed sales mix. It does not prove that scheduling causes the country differences. Nor does a smaller live share mean live is unprofitable. Without costs, margins and comparable conversion data, calling the biggest sales format the most profitable would be a step too far.
Why it matters: Start a format plan with how customers actually buy, then test what the business can deliver profitably. These estimates cover four markets, not all Europe or the world, and offer no trend over time. The broader lesson for sellers is to match content production to the local buying route and measure the result. A busy live studio is an activity; profitable orders are the objective.
Sources: ECDB sales-format analysis and its original post, using Kalodata and ECDB estimates for 2025.
