metrics block measured over the period you request.
Shop
A storefront on a channel — the seller, not the brand. One brand can operate several shops across markets, and each is a separate entity with its own ID. Attributes —name, handle, url, logo_url, seller_type, categories, rating,
reviews_count, products_count, opened_at.
seller_type normalizes each channel’s own seller taxonomy into brand, reseller,
cross_border, local or unknown. It is the fastest way to separate first-party brands from
resellers in a result set.
Metrics — revenue, units_sold, avg_unit_price, revenue_growth_rate,
revenue_by_source, products_sold_count, creators_count, videos_count, livestreams_count.
revenue_by_source is the shop’s revenue split across selling surfaces, as ratios summing to 1:
Product
A listing sold by one shop in one market. The same physical item sold by two shops is two products, because price, commission and performance all differ. Attributes —title, url, image_url, brand, shop, category, price,
commission_rate, rating, reviews_count, listed_at.
price carries current, min and max — the observed range over the period, which is how you
detect discount-driven spikes.
Metrics — revenue, units_sold, avg_sale_price, revenue_growth_rate,
revenue_by_source, creators_count, videos_count, livestreams_count, gpm.
Creator
Someone who drives attributed sales. Ranked by what they sell, not by audience size. Attributes —name, handle, url, avatar_url, bio, followers, categories,
agency.
categories is ordered by the creator’s revenue share, not by self-declared niche. A creator who
posts about fitness but sells supplements shows up under supplements.
Metrics — revenue, units_sold, gpm, avg_commission_rate, products_count,
shops_count, videos_count, livestreams_count, avg_views_per_video, engagement_rate,
follower_growth.
Video
Short-form content with products attached. Attributes —caption, url, cover_url, duration_seconds, published_at, creator,
products.
Metrics — views, likes, comments, shares, engagement_rate, revenue, units_sold,
gpm.
Video metrics are lifetime-to-date rather than period-windowed: a video’s view count is the total
it has accumulated. The period on a video search filters which videos are returned by
publish date; it does not slice their metrics.
Livestream
One live selling session, from start to end. Attributes —title, url, cover_url, started_at, ended_at, duration_seconds,
creator, shop, products.
Metrics — revenue, units_sold, viewers, peak_concurrent_viewers,
avg_watch_seconds, likes, comments, gpm.
Sessions appear once they end. A session in progress is not returned, because its metrics would
change under you between pages of the same result set.
Category
A node in the normalized category tree. Category IDs are channel-independent — this is the one entity whose IDs are shared rather than per-channel, which is what makes market sizing portable. Attributes —name, parent_id, level, path, children.
Metrics — revenue, units_sold, avg_unit_price, revenue_growth_rate, shops_count,
products_count, creators_count.
Because
cat_ IDs are shared, a filter written as "category_ids": ["cat_beauty_personal_care"]
keeps selecting beauty when you enable a second channel — even though that channel’s own
taxonomy is shaped differently.References between entities
Nested entities are returned as compact references — anid plus enough to display it — never as
full records. A product carries a shop reference; a video carries a creator reference and a
list of products.