A week ago I attended a customer briefing here at the Singapore Executive Briefing Center (EBC). One of my colleagues gave an overview on business continuity and disaster recovery (BCDR). His presentation included a number that I posted on Twitter. David Manconi asked for supporting evidence of my claim so I thought I would post it here.
First, here is a grab from the slide I shared to the Twitterverse.
In summary of this slide, I tweeted, “The average cost of downtime in a disaster is $223k/hr. Disasters: expensive. #VMware #SRM: cheap. http://pic.twitter.com/AAbIrUMS”. David saw my tweet and asked for external support for this claim. I searched, and found nothing from the claims author (IDC/Alinean).
I did, however, find a another vendor with the same data. The presentation notes in my colleague’s presentation said the $223,000 number was an average from a range including costs from $42,000 to $600,000. By searching for these numbers I stumbled across a conference session with more detail. In this second presentation, apparently delivered by SAP at a CA conference, cost of downtime is reported by application type. Here is their slide.
So, it seems EMC’s $223,000/hour number was calculated from the range above. A flat, basic, stupid average from the above table calculates a cost of downtime of about $193,000/hour. Finding the average cost by taking the average of those numbers is dumb, of course. But we could weight these application types in a hundred different ways. I feel fairly confident that somebody tried to collapse all eight numbers into a single, monolithic assessment that works in marketing slides. With slight weighting, they could have produced a $223,000/hour number.
So, to David and the rest of my readers, I do not have a public link from Alinean that I can point you to. I suspect that Alinean provided EMC and SAP these numbers for us to use with our customers. I conclude that we can trust the cost of downtime to be near $223,000/hour to the extent that we can trust Alinean.