A recently published academic paper reports an analysis of profiles on AW (
External Link/Members Only), unfortunately their lack of knowledge about punting and AW makes some of the conclusions either blindingly obvious or a bit bizarre.
What the authors say in the abstract is “This study investigates the determinants of pricing and popularity in the market for commercial sexual services online by using data from the largest UK network of online sexual services, a platform that is the "industry-standard" for sex workers. While the size of these influences vary across genders, nationality, age, and services provided are shown to be primary drivers”
They “scrape” data from AW profiles-their first error is to believe that male advertisers are genuine rather than chancers hoping for a free shag. You can see this as (a) those advertising as bi-or straight have virtually no reviews and (b) there is virtually zero market for men advertising to women. Gay men tend to advertise on specialist sites and not AW.
Then they take things such as advertised age as genuine-as well as dress size etc and try to correlate this with number of profile views or number of positive reviews..
From my point of view one of the more interesting findings is that they use a statistical clustering technique to group which services are advertised together-they identify five groupings-one contains services that most advertisers provide but there are other clusters or instance based around duos or moresomes or around humiliation domination.
Overall their lack of knowledge about how AW or punting operates shines through. It’s good fun to pick holes in the paper. But data scraping of this sort could be used for more interesting things-such as which phone numbers tour together-which appear attached to different profiles over time. Also what would be really useful is trying to search to see if photogalleries are nicked from elsewhere