Furthermore, "gentleman's agreements" like these are fundamentally anti-competitive.
That's quite a stretch, given the usual definition of anti-competitive. A "gentleman's agreement" to not lower prices, or not to expand into a certain market, etc., would be genuinely anti-competitive. This is simply an agreement to not kick somebody when they're down. That's just good sportsmanship and integrity. Personally, I think more companies should display this kind of thinking.
Non-explicit agreements become popular when contracts are unenforceable. They also confer an advantage to groups that are more likable. Does that mean that only 'unlikeable' groups would oppose gentleman's agreements? Not really, because they also automatically favor incumbent firms over early stage firms or anyone who is disruptive.
The nod and the wink about refusing to compete, and advocating an effective blacklist of companies that evangelize on the subject of reliability can't be anything more than a nod and a wink, because it would put the signatories in legal jeopardy.
Alternately, they could build extra capacity, and sell terms on how often they would provide overflow capacity to each other, but that would cost more than simply campaigning to prohibit discussion of failures.
There really isn't any kind of stretch to call this anti-competitive. It is pretty much the archetype for using an informal agreement as an enforcement mechanism in order to avoid a mutually detrimental nash equilibrium. More companies could engage in the type of behavior advocated here, but then it wouldn't be the dynamic industry that it is.
This resembles a statement to the effect of not considering an economic phenomenon because it was not immediately obvious.
Professional courtesy occurs in many professions including law, medicine and finance, and it causes real economic harm even though the harm is so spread out that it frequently goes unnoticed. The very nature of this type of collusion is that it is not obvious on the surface, and though the harm is real it is often well-intentioned and understood as courteous by the people responsible for it.
Only marginally more obvious is if you even hear people describing whether they raise the hourly rate for their gardener as a courtesy issue in regard to a friend who employs the same person. There the harm is almost blatantly obvious, but I assure you it is a common sentiment.
Here's an actual quote:
>When someone you compete against is suffering, especially as a result of any kind of infrastructure issues, shut up and keep your head down.
and, later:
>Not to mention, do you really want those customers who were so quick to leave? What do you think they’ll do when you go down next time?
which sounds like the argument about trying to hire A, B or C people—it's actually a pretty good idea to target really demanding customers. There is a random component to infrastructure failures, but it is nonsense to pretend that preparation, capacity, and redundancy have no effect, and that they are not only effective marketing topics, but features that are ultimately very important to users.
it is nonsense to pretend that preparation, capacity, and redundancy have no effect, and that they are not only effective marketing topics, but features that are ultimately very important to users.
FWIW, I never claimed that, and I don't think the OP did either. At least for myself, I am very specifically referring to the "black swan event" situation, where something external, unlikely and damaging happens to a competitor. In that case, I don't see how showing some restraint, and not launching a targeted campaign to try and leech their customers, is anti-competitive. And to the extent that you can, somehow, define it so, I would argue that any damage is so absurdly trivial that it can be ignored.
When Rackspace or Amazon Web Services go down, customers recognize that they are fragile, and no one claims that black swan events are irrelevant. The discussion around black swan events revolves around how bad our instincts are when it comes to the high likelihood of rare events. And, the point is that they should not be excluded from our models.
If the two hurricanes and two tsunamis that rose to the level of 'astonishing' during the past decade are the only time that people grasp their likelihood, it is counter productive to treat these events as irrelevant. The same goes for DDoS attacks.
It doesn't matter what people think about this specific issue, since there should be no doubt that there will be many more attempts at this type of vigilantism. No one wants their company beholden to other people's whims and angers, so they should have better defenses. Whether competitors point out their failings or not should be the least of anyone's worries. The capacity to handle attacks up to varying levels of size shouldn't be an impolite or unmentionable subject, it should be a key part of business contingency plans for businesses and their customers.
That's quite a stretch, given the usual definition of anti-competitive. A "gentleman's agreement" to not lower prices, or not to expand into a certain market, etc., would be genuinely anti-competitive. This is simply an agreement to not kick somebody when they're down. That's just good sportsmanship and integrity. Personally, I think more companies should display this kind of thinking.