This post originated from an RSS feed registered with Java Buzz
by Michael Cote.
Original Post: PeopleOverProcess.com: SaaSy Downtime
Feed Title: Cote's Weblog: Coding, Austin, etc.
Feed URL: https://cote.io/feed/
Feed Description: Using Java to get to the ideal state.
One of the stupidest, yet long-lived, ad campaigns is the old cable industry one about satellite TV: you know, there's a thunderstorm coming, so Dad gets up on the roof, and gets hit by lightening, and then his whole family ends up on the dole, selling their kidneys to make the mortgage before having to sell their children into white-slavery for the SUV payments.
Man. Shoulda gotten cable.
Uptime
The absurdity of the campaign doesn't come from it being false, the sillyness comes from the fact that every system experiences downtime. Several years ago, while I was still in bachlorhood, I subscribed to a DVR service from my local cable provider. Instead of TiVO, it was a Scientific-Atlanta box and TimeWarner Cable.
Let me tell you this: that thing was always busting. I called the cable company once or twice, and after 40-60 minutes on the phone, the solution was always the same: reboot it and kill a chicken...maybe it'll work.
It was like the Windows 95 of TV. And, being in an apartment, the cable and broadband connection would get spotty if my neighbors were downloading too many of those hilarious monkey videos.
The point is, every system screws up in it's own way. Hosted applications will be down at times, desktop applications will freeze, or, worse, zap all your data when your hard-drive gives out. In my mind, quantitatively, all systems fail the same amount of time: it's just the quality, so to speak, of those failures that's different.
What matters is what your vendor (or service provider in the case of cable) does when it breaks.
"It may be down, but they're so nice!"
I'm almost tempted to say that how nice and open folks are while they're fixing your problems is even more important than how quickly they fix it. As an example, my feed reading service, FeedLounge went down yesterday for about half the day. Normally, I'd be frustrated as hell waiting for a service as critical to my life as FeedLounge is to come back up. But, somewhat ironically, because it'd gone down before (a few times even), I knew it would be up again soon enough. More importantly, they've setup a whole site to track uptime and problems: the brilliantly named FeedLounge.info.
If you fear telling your users your real uptime, think of it this way: once you report it, you control the story of your uptime. It's hard to start up a scandal around facts that companies are open and truthful about it; which explains one reason why politics are so rife with bickering and innuendo.
This is just the SaaS version of what we call "customer service," which the IT world is terrible at. Take a gander at Lenovo's David Churbuck's post on the topic for a thorough discussion. Indeed, stealer support and, it's sister, dependability are the core components of ThinkPad lore. Only AppleCare comes close in my mind. Dell? Just trash it and buy a new one; they're priced accordingly.
The whole time, The Pilot gave us frank, truthful updates along the lines of, "I have no idea what’s happening. They said the [rescue] plane would be here at 7, and it’s not. I’m sorry. I don’t know what’s going on."
Or, "I know how you feel. The crew and I have been up since 5AM."
The Pilot didn’t bull-shit us, or even put on a "everything is fine" face. Meanwhile, he was constantly on the phone to "The Company," it seemed, wheeling and dealing on our behalf. I over-heard someone saying that jetBlue doesn’t have a large customer service department: The Pilot and crew are the customer service department.
Where ever you find truthful, human interactions when the Shinola hits the fan, you'll find loyal customers and stable revenue streams.
Disclaimer: as James said, "Lenovo is not a client, but we'd like it to become one." Though, it's just happy coincidence that we talked with David yesterday and, thus, that this post laces in ThinkPads. It was really FeedLounge's downtime that prompted this post.
Previous Topic
