Reading seller weight without a scoreboard
A handful of publishers can fill a category's sight-lines. That is not depth. It is a chorus.
Seller weight is a blunt phrase for a simple picture: how much of the visible aisle belongs to a few publishers. If three houses occupy the featured row and refresh daily, the category can look thick while new publishers never receive a first conversation.
We mark this on paper. We do not need a live scoreboard. A table of publisher identifiers against listings and featured days is enough. The room can see whether “two hundred apps” is two hundred voices or twelve voices in many costumes.
Bilingual catalogs in Malaysia make the costume easy. The same tutoring app appears once in English and once in Bahasa, sometimes in two categories. Counted naively, that is depth. Counted as a pair, it is one stall with two signs.
Operators sometimes want a single concentration number they can take to a board. We will give a share if the sample supports it. We will not invent an index that pretends every category is comparable. Tutoring and tow trucks do not share a natural number of publishers.
The useful argument in the working session is not “is concentration bad.” It is “does this aisle still have a walker who will try an unfamiliar publisher.” If the answer is no, featured policy is doing the teaching, not the rest of the shelf.