Only Polly Pocket began because a family had a pile of tiny dolls and playsets and could no longer remember which little plastic person belonged in which little plastic house.
The site’s own history says it started in 2000 primarily to identify sets owned by the family’s children. Collectors began using it, visitors asked whether sets were for sale, and the project grew into a reference site for vintage Polly Pocket toys.
The result is the sort of database that looks simple until you realize how much hand sorting sits underneath it.
Tiny toys, large indexing problem
The main navigation runs through production years from 1989 to 2002, with additional sections for Disney sets, undated items, characters, and related treasures.
Individual pages help identify sets by year, name, dolls, and accessories. That matters because vintage Polly Pocket collecting is unusually vulnerable to mixed parts. A compact may survive while its original figures disappear, migrate into another set, or get replaced with something that merely looks plausible.
The site approaches that problem visually and chronologically.
It also preserves the social layer around collecting. Character pages introduce Polly’s friends; collector links point outward; and the archive gradually became substantial enough that its creator later produced a 504-page printed collecting guide based on the same body of work.
The interesting part is how ordinary the origin was.
Nobody started with a plan to build infrastructure for a collector community. They made a family reference page because the toys were confusing. Enough other people had the same problem that the private inventory became public knowledge.
That is one of the most reliable patterns on the old web.
CacheRat’s 1,967 Ancient Web Domains research list contains many sites where a personal catalog quietly became a specialist reference because the owner kept adding details that larger databases ignored.
Only Polly Pocket is a particularly cheerful example.
The objects are tiny.
The taxonomy is not.