Anyone buying into a Web-based service knows about the SLA - the service level agreement. That's where the Web company makes a promise about uptime, the amount of time that the service will be up and running without any service disruption.
In most cases, there's a clause in the agreement that allows for scheduled downtime for maintenance. Now, Google - in an effort to further set itself apart from competitors - is removing that downtime clause from its customers SLA's.
From here on out, any downtime will be counted and applied toward the customer's SLA. In addition, the company is amending the SLA so that any intermittent downtime is counted, as well, eliminating the previous provision that any downtime less than 10 minutes was not counted. In a blog post, Google Enterprise Product Management Director Matthew Glotzbach wrote:
People expect email to be as reliable as their phone’s dial tone, and our goal is to deliver that kind of always-on availability with our applications... In 2010, Gmail was available 99.984 percent of the time, for both business and consumer users. 99.984 percent translates to seven minutes of downtime per month over the last year. That seven-minute average represents the accumulation of small delays of a few seconds, and most people experienced no issues at all.
And, for those wondering how the downtime compares to on-premise email - specifically for Exchange customers, Google says that seven minutes compares "very favorably." The company cites a 2010 report from the Radicati Group to conclude that that suggests that Gmail is 46 times more available than Exchange.
Google said the Radicati report only measures uptime for on-premises Exchange, Groupwise and Lotus deployments but does not measure performance of hosted email systems.
Google said that data about Microsoft's BPOS cloud offering was unavailable but that its own analysts of Microsoft's service notifications showed 113 incidents last year, with 74 planned outages and 33 days with planned downtime.
updated: I reached out to Microsoft for a response to Google's calculations Friday morning but didn't hear back until late Friday. The company neither confirmed nor disputed the numbers but instead sent a statement that wasn't attributed to anyone in particular at the company. That statement read:
Microsoft Online Services offer the industry’s most rigorous financially-backed SLAs. We guarantee 99.9% uptime, or we give customers money back. We’ve always counted any service issue as downtime, from the minute it starts to resolution. We also count issues for any number of impacted users, not just if “enough” users are impacted. What’s more, Microsoft offers 24/7 customer support regardless of the level of impact. We know our customers expect guarantees they can trust, and that’s what we deliver. Microsoft Online Services have averaged 99.9% or better uptime for the past year.