Diseconomies of scale in fraud, spam, support, and moderation

Re:Diseconomies of scale in fraud, spam, support, and modera... by faronel on Monday February 19, 2024 @01:47AM (#39426826)

From the article as an example of what doesn't scale:

"...He was going through emails in his inbox, then responding to questions in the craigslist forums, and hopping onto his cellphone about once every ten minutes. Calls were quick and to the point "Hi, this is Craig Newmark from craigslist.org. We are having problems with a customer of your ISP and would like to discuss how we can remedy their bad behavior in our real estate forums". He was literally chasing down forum spammers one by one, sometimes taking five minutes per problem..."

Seems like it's possible but not feasible to build a pipeline to scale out that sort of behavior.

[ Parent | on HN ]

Re:Diseconomies of scale in fraud, spam, support, and modera... by saagarjha on Monday February 19, 2024 @04:43AM (#39427781)

Why not? It doesn't have to the founder on every call if you are looking to scale.

[ Parent | on HN ]

Re:Diseconomies of scale in fraud, spam, support, and modera... by michaelt on Monday February 19, 2024 @05:45AM (#39428228)

It depends if your revenue scales as fast as abuse scales.

Amazon should be able to keep on top of fraud easily, twice the sales means twice the revenue means they can afford twice the fraud checks, if they want to. That fake $30 "2tb microsd card" puts $30 of real money in their bank account.

But for something like Twitter, revenue is only vaguely related to number of tweets. That "elon musk wants to give you free dogecoin" tweet doesn't make them a cent.

[ Parent | on HN ]

Re:Diseconomies of scale in fraud, spam, support, and modera... by Jochim on Monday February 19, 2024 @10:07AM (#39430532)

Then their business isn't viable.

The alternative is that we allow businesses to externalise those costs onto society.

[ Parent | on HN ]