EBay is cutting about 1,000 roles, its second round of layoffs this year. Two rounds in twelve months is not a cost program, it is an admission that the operating model needs a different shape. For a marketplace that once defined online resale, the question is no longer how to grow faster than rivals. It is how to stay relevant when discovery moves to assistants and when new entrants own the categories that used to be its core.
A second round in one year is a pattern
First rounds are usually about removing duplication after an acquisition or trimming a pandemic era build out. Second rounds are about structure. When a company returns to the same lever within a year, the underlying revenue per employee has not recovered, which means the problem sits in the mix of categories and the take rate rather than in overhead. EBay layoffs
The timing is also telling. The cut lands just before the peak season, when marketplaces normally add support capacity rather than remove it. Management is signaling that it would rather protect margin through the peak than chase volume. That is a rational choice for a mature platform, and a difficult one for sellers who depend on support responsiveness in November.
The marketplace model is being repriced
Marketplaces earn from three sources: take rate on transactions, advertising, and services to sellers. All three are under pressure at once. Take rate faces resistance because sellers can compare fees across more channels than ever. Advertising faces resistance because retail media inventory has multiplied. Services face resistance because third party tools now do what platforms used to bundle. retail sales data
The category mix compounds the problem. Higher prices pushed September retail sales up 6.7 percent over last year, but the increase came largely from price rather than units. On a marketplace, a price driven sales number with flat units means the same gross merchandise value generates less shipping volume, less ancillary fee revenue and more price sensitivity from buyers.
Where the cost actually sits
Take rate pressure
Sellers evaluate a channel on total cost of sale, not on commission alone. When fulfilment, returns and advertising are added, the effective take rate on some categories approaches the point where a direct to consumer store becomes cheaper. Platforms respond with fee changes, but every change redistributes margin between categories and invites sellers to shift assortment elsewhere. Adidas Q3 results
Category mix
Resale, collectibles and parts have loyal buyers and low return rates, which is why they survive fee pressure. Fashion and electronics have higher return rates and more price transparency, which makes them expensive to host. A marketplace that cannot choose its mix ends up subsidizing the expensive categories with the durable ones, and that subsidy eventually shows up as layoffs.
Buyer acquisition
Acquiring a buyer is more expensive when discovery fragments. Search, social, video and now assistants all compete for the same attention, and each channel requires its own creative and its own measurement. For a platform without a membership flywheel, that cost recurs every season rather than compounding into loyalty.
Adidas and Nike show the same split
Adidas posted a record third quarter while Nike continues to work through an uneven turnaround, which is the same story at brand level. Execution and category positioning decide outcomes more than the macro environment does. A platform facing structural pressure cannot blame the cycle alone, because competitors in the same cycle are growing.
For sellers, that means channel concentration is now a real risk. A marketplace that is restructuring twice a year will change fee structures, support levels and promotional terms. Sellers who have built their entire economics around one platform are exposed to decisions they do not control. Diversification is expensive, but so is dependence.
What sellers should plan for
Expect three things over the next two quarters: fee rebalancing toward categories the platform wants to keep, tighter enforcement of listing quality, and reduced human support in favor of automated resolution. The third is the one that hurts most in peak season, because automated resolution handles standard cases well and unusual cases badly.
The practical response is to reduce the number of unusual cases. Clean listings, accurate attributes and clear return policies cut the volume of exceptions. Sellers who invest in that discipline will find the automated systems work in their favor. Sellers who do not will experience the restructuring as a service outage rather than a cost change.
