My fellow Irregulars Dan Farber, Charlie Wood and Charles Zedlewski blog about salesforce's planned outage of 6 hours.
It's good to hold salesforce to a high standard. But let's seek the same benchmark of SAP, Oracle, Lawson, Microsoft, and every other major on-premise vendor.
Let's ask them:
a) what percentage of your customer base expects to be up at least 100 hours a week,
out of the possible 168?
b) how much planned and unplanned downtime is there each month is in each of your customers?
c) what is it costing the customer base to host, per user, the application and related data?
d) what SLAs do they have with their outsourcer?
e) how many times has that service level been breached?
If we can get reliable data from these vendors, we will likely find that in spite of talk of "mission critical" applications and needing premium priced, kevlar coated data centers, less than 25% of their customers need the systems to be up more than 100 hours a week. And if they would re-bid their hosting services using the scale, location and efficiency of new data centers being built by salesforce, Google and other large SaaS vendors, their customers are over paying for incumbent hosting services by a factor of at least 3X.
But don't take my word for it. Let's get data from the on-premise vendors. Then we can pick on all of them where they need to improve - not just salesforce.com.
