The simple answer is – no. That was easy, let’s move on to the next topic!
Okay, so let’s examine the proposition in a little bit more detail.
It’s all but impossible to make the necessary financial calculations to work out the exact cost of downtime to any particular business, town, city, region or country. It doesn’t stop folks trying, though! Various suggestions as to the cost of downtime to a business include $100,000 an hour, more than $300,000 per hour, or even over $1 million an hour. If we might borrow from Donald Rumsfeld for a minute, the main problem with trying to come up with an accurate figure as to the cost of downtime revolves around trying to bring together the financial impact of the known known costs, the known unknown costs, and the unknown unknown costs!
In other words, it’s relatively easy to work out how much you pay your staff and, therefore, how much money you ‘lose’ if they have to sit idle for a couple of hours – although it’s more than possible that they can do other tasks that don’t rely on the IT function. At the other end of the scale, how can any business ever calculate the cost of losing a potential customer? The XYZ company is about to sign up with an MSP (unknown to the MSP at that stage). The MSP suffers a well-publicised outage, and the XYZ company changes its mind and goes elsewhere.
I wouldn’t discourage anyone from calculating a few sums as to the obvious financial impact of any power outage on their business, but rather more important is to understand the overall impact of a power failure. For a Cloud/Colo/Managed Services provider, there’s the obvious compensation offered to existing clients. However, where customers might once have been happy to, say, receive a month’s power for free by way of an apology, service level agreements are becoming more and more involved, to try and reflect the true impact of IT downtime on a customer’s business. Our XYZ company sells £50,000 worth of goods an hour from its website, the website is offline for two hours, thanks to a power failure at the provider, and the compensation bill is the sum of the two: £100,000. Ah, but no provider is going to agree to such levels of compensation that will put it out of business. And here we enter the murky world of cyber insurance (a developing industry) and the give and take between provider and customer. It probably doesn’t make sense for the customers to insist on compensation that puts their provider out of business, but, then again, these customers have lost significant revenue, and suffered reputational damage as well.
Likewise, the provider has suffered a damaged reputation. When existing clients’ contracts come up for renewal, they might decide to head somewhere else and, as mentioned previously, what about all the potential customers that are put off by the adverse publicity surrounding the provider’s power failure?
The solution? Well, in simple terms, there are two types of Cloud/Colo/Managed Services providers out there right now. There’s the ‘pile it high, sell it cheap’ folks, who will almost certainly respond to any downtime issue with a variation of the response: “Well, what did you think you were getting for that price?” And then there are the providers who are genuinely interested in building a partnership with their customers – and who will go that extra mile to try and ensure that there is no downtime, but also work with their customers to resolve any service issues and quickly and amicably as possible.
The whole area of third party data centre and IT provision is relatively new in its present form (ie where IT is a critical part of so many businesses), and there’s still a long way to go before any standards or best practices are agreed upon and published.
The bad news is that this means there’s a great deal of confusion and uncertainty as to how to resolve any downtime/service issues. The good news is that service providers have the opportunity to create for themselves a great reputation for resolving such issues speedily and to the satisfaction of their clients.
Service providers also need to decide at which end of the market they wish to operate: low cost services, with relatively low levels of customer support; or, the high end of the market, where the cost of services reflects the customer support that is provided as part of the overall package.
As the mobile phone industry continues to demonstrate (at least in the UK), offering a low cost service, with virtually no customer service, in the knowledge that you are going to lose a load of customers, but also pick up plenty of new ones from your competitors, is a perfectly viable business model. However, do bear in mind that data centres and IT are normally somewhat more business-critical than a mobile phone, so customers will be rather more stressed, and expecting a better level of service, when things do go wrong.
Finally, barring truly extraordinary sets of circumstances in at least two different locations, power outages shouldn’t really have an impact on Cloud, Colo or Managed Services provision. While there might be a significant cost associated with ensuring such resilience, the alternative - taking a risk with your power infrastructure, and potentially suffering downtime, could just spell the end for your business.
What sort of service provider do you want to be?