Greg Linden wonders if there's a tendency to lower uptime expectations among web companies, and document recent examples :
Google's Blogger has been taking many planned and unplanned outages over the last few weeks. Yahoo My Web just announced they're going down "for a few hours." Salesforce.com just had multiple outages, including one that lasted almost a day. Gap decided to close Gap.com, OldNavy.com, and BananaRepublic.com for over two weeks. Bloglines took an outage instead of trying to switch to a new data center without any interruption of service. Technorati did something similar a year ago, but took a longer weekend outage for the move. And there are many, many other examples.
He's also fingered David Heinemeier Hansson's opinion piece, Don't Scale, with The folly of ignoring scaling, and points to other interesting opinions: Om Malik's post, "The Web 2.0 hit by outages" and Nicholas Carr's post, "Salesforce.com's hiccups".
Lowering uptime expectations is to me akin to Microsoft getting the whole world think that it's normal for an application to crash several times a day and your computer to show a blue screen of death once in a while ("What, you didn't save your work? Your bad, dear!")
Trouble is, when your software set is what it is because of tradition, pressure, inertia, *cough* monopoly *cough*, and it costs thousands to change it, you think twice before dumping it. But in the Software as a Service (SaaS) world, where the "no strings attached" label is touted along a low monthly price in a much more competitive space, this is a very different story.
Still, I believe it's far more difficult to switch SaaS suppliers than, say, more commoditized ones such as ISPs. Internet connectivity comes in the same flavor everywhere, but when you've got sites and mailboxes, it starts to become a little more complicated. And when you've built critical business functions around a platform such as salesforce.com, where do you turn to and how long/difficult/costly is it if you want to drop them?
I've now spent an equal part of my professional life on both sides of the client/supplier frontier, and I've learned how crucial — and mutually beneficial — it is to build trust. Nicholas Carr is spot on, citing David Berlind on Saas reputation about overpromising reliability and underdelivering, in that SaaS providers need to be totally transparent and realistic in order for this model to fly:
No system is 100% reliable, and I believe the SaaS model will ultimately prove to be considerably more reliable than the on-premise model, but obfuscation about an outage may end up doing more harm to a SaaS provider than the outage itself.
On a side note, and reading Greg's post on Blogger and database inconsistency, I'm afraid that there's already a two-class model, where the B2C SaaS providers may show far less consideration for their customers data than the B2B ones. As our life goes more and more digital and online, that's not good. As a matter of fact, the dreadful shrink-wrap license of software that would exonerate the software vendor from all responsibilities, has turned into a simple "accept it or go away" checkbox on all those SaaS sites, and you'd better read the fine prints very, very carefully before subscribing. More transparency would be appreciated here too, for example on the fact that if your supplier boasts its duty to keep your data safe as a key advantage of SaaS vs. on-premises backups, what use is it if the contract explicitly stipulates that you cannot claim any damages from data loss? Clearly, when Six Apart's Typepad service went South, their customers' reactions did show a high level of reliance on the platform (a number of them being businesses!) but also a glaring misconception about the legal responsibilities of such a supplier for this kind of service, and what they could really claim in case of an outage (answer: nothing). I wonder if the insurance world has started to integrate SaaS in their coverage of business continuity. This would be another strong positive signal for this business model.