I remember reader Ajay wondering how those godawful “Aberdeen Steak House” things around the West End have a business. I can’t find the discussion now, but I recall I told him that their ideal customer was someone for whom paying over the odds for a really bad dinner was an important part of their night out. That’s how they knew they were having a good time, I speculated.
Now here’s a Microsoft Research paper that explains this more elegantly. It examines why 419 spammers are so obvious and their production values are so crappy. Basically, the problem is false positives – it’s very important to the spammer to target suckers and to avoid wasting effort on non-suckers, and because the pool of potential marks is big relative to the pool of suckers, anything that improves the targeting is a disproportionate boost to the spammer’s payoff.
Even though the cost of sending out spam is minimal, this is only the first step in the process – once a mark responds, the attacker starts to incur costs. Non-suckers who respond will get wise at some point, leaving the attacker with a loss. Because suckers are rare, it’s hard to find a way to predict who might be a sucker. So the optimal strategy is to broadcast as widely as possible, but to tailor the message. The reason why they look like only a real sucker wouldn’t spot them is that they’re specifically designed to be easy to spot, so as to put off the non-suckers. An upshot of this is that the people who get their kicks by stringing 419 spammers along may actually be doing useful work.
I suspect this phenomenon – essentially the opposite of advertising, a sort of negative marketing – is much more common than we may think, and that it explains much else beyond terrible restaurants. Boris Johnson comes to mind, as do quite a few other politicians. In a low-turnout context, it ought to work, especially if you can put off the non-suckers from voting at all.