I think we are seeing a later day MMORPG forum. At some point, it is easier to argue for a patch for your class than get 1 pct more stats, so you stop playing the game and start playing the boards. ZenDesk is the vendor most likely to drop prices by 1k for a single fauxraged posting, so start your keyboards.
However I think the biggest thing is they didn't sign up for $1,000. They signed up for half that. Regardless of the figures no one likes seeing a 74% raise in cost.
And there's nothing wrong with raising your price by a large amount, even if people are locked in with your product - you simply grandfather existing users.
You accomplish what you want, and your existing users win too because all of a sudden they're getting a great deal.
Nobody leaves. Nobody complains. Everyone's happy.
For example, even though Microsoft Office probably provides a lot more value than what it costs, it they doubled their prices overnight, would you not expect any outrage?
This breaks down severely on these web models, where you're more inclined to start out with a lower price and feature set while you gain customers, but want to increase the price as you add features (and value)
Its sticky and makes setting your first price point a more critical game.
I wonder if people have thought about "versioning" their SAAS apps. (do they? I don't have a ton of familiarity here). Personally I think Zendesk has to reward their early customers by keeping them at their current rate.
Short summary: anchor points play a great role in setting human expectations. When plaintiffs ask for exorbitant amount of money*, then usually will get more money than they really deserve because jury's anchor point is set high.
The definition of an "agent" is pretty vague for many places too. We, for instance, could probably have 40 "agents" handling help desk duties if we let our volunteer forum moderators help out with user support issues like I'd like to.