PPLOTTDATA
PLOTT DATA · marketplace growth brief for Amazon11 marketplaces · 2026-10

Which categories grow, how large they are, and how fast they grow.

We tested eleven marketplaces. We tested each one live. For each marketplace, this page shows three things: the data we measure today, the data we estimate, and the history that exists. Trading cards answer all three questions with real sales data. The other marketplaces give measured supply and an estimated demand. We mark every field.

Marketplaces tested
11
Resale, live commerce, general, and specialty.
Answer size today
11 / 1
Supply on all. Demand on one.
Carry real history
5.3yr
Trading cards: weekly units and prices from June 2021.
Days to a growth signal
4wk
Review speed and rank changes become stable at week four.
01
The three questions

One question. Three different tasks.

The first task is to find the categories. This task is easy. Almost every marketplace lists its categories. The second task is to measure size. Supply is public. Demand is often not public. The third task is to measure growth. Growth needs history. Almost no marketplace gives this history. We answer each task for each marketplace.

Find the categories

Measured

This task is complete for the whole set. Each marketplace lists its own categories. We map all eleven marketplaces to Amazon categories. A direct comparison then becomes possible.

Measure the size

Partial

Supply is the count of sellers, products, and prices. We measure supply almost everywhere. Demand is the count of units and the sale value. One marketplace makes demand public. We estimate demand for the others, and we mark each number.

Measure the growth

Estimated

No party can add past history. History starts on the first day of collection. Trading cards include history. For the other marketplaces, we start to record now. A growth rate becomes reliable at about week four.

02
Coverage map

Eleven marketplaces, by what they answer.

Tier 1 answers two or more of the three questions today. Tier 2 gives the supply and depth layer. Each row is a surface that we tested live.

Marketplace
Categories
Size
Demand
History
TCGplayerTrading cards
meas
meas
meas
meas
WhatnotLive commerce
meas
part
est
part
eBayGeneral
meas
meas
est
part
PoshmarkResale fashion
meas
meas
est
part
DiscogsMusic and media
meas
meas
est
part
GOATSneakers
meas
meas
est
part
Back MarketRefurbished electronics
meas
meas
est
part
Walmart MarketplaceGeneral
meas
meas
est
part
MercariResale
meas
meas
est
part
StockXSneakers and collectibles
part
part
est
est
SHEINFast fashion
part
part
est
part
Measured read from the marketplacePartial partial coverageEstimated from a named method
03
The rule

Measured data and estimated data stay separate.

A measured field comes from the marketplace. An estimated field comes from a named method. We do not mix the two. Every number carries its label.

TCGplayer

Trading cards

Each product has a daily and weekly record of units sold, average sale price, and transaction count. This is the one marketplace that publishes real demand.

Whatnot

Live commerce

A category tree of 35 nodes. A seller ranking for each category over three time windows. Public seller totals. The full category sale value opens with a partner login.

eBay

General

Live listing counts and full search results across 1,454 category nodes. The sold-price records need an account login.

Poshmark

Resale fashion

A full category tree and brand-level listing counts. Likes, shares, and offers give an interest signal.

04
Where to start

Two marketplaces answer the question from end to end.

Trading cards is a category that Amazon competes in. It is the one marketplace with public unit and price records, and it has five years of history. Whatnot is the live-commerce surface and the marketplace in the brief. Its categories and its seller activity by category are available today.

05
Next step

What we build first.

We build a daily snapshot store across the Tier 1 marketplaces. We date each record from the first day. We add a weekly summary. This is the same system that already runs in a Wayfair pilot. In four weeks it gives rank, review, and price change for every category. One quarter of client unit sales then calibrates the estimate into a unit model.

Day 1
Full cross-section. Category tree, every ranked product, supply count, price, rating, and review count.
Week 1 to 2
First real movement. Rank changes, price events, new products, and products that leave.
Week 4 and after
Review change gives enough signal to estimate units. The link between rank and speed becomes stable.
Quarter 1 and after
Season and promotion effects. A calibrated unit model across competitor products.