I’m sure that you’re aware that application downtime costs a boatload of money. But let me start this blog by reinforcing something you already know. Aberdeen Group estimates that application downtime costs $110k per hour for an average company. Ponemon Institute claims application downtime costs $205k per hour and Gartner reports that it’s $42k per hour. Pick a number – they’re all big and significant. But this isn’t the whole story. The potential cost of application downtime is actually much, much greater. Let me explain.
The fact that so much money can be lost during application downtime results in a myriad of uptime monitoring solutions and application performance management tools. Companies like CA Technologies provide such solutions and there is no doubt that these solutions are of great value. However, even after you implement these tools you’re still going down. Machines break and humans make mistakes and there is nothing you can do about it. We live in an imperfect world. There is no magic pill that you can take that will ensure that you never have an outage.
As a case in point, look at Yahoo Mail. As recently as December 2013 they had an epic fail. A Twitter user @wonder_wendy, tweeted, “Very frustrated!!! My account has been down since Monday. Apparently due to “maintenance”, with no heads up. Terrible!” This my friends, sums up the hidden cost of application downtime. We all know about the costs reported by Gartner, Aberdeen and others. What we neglect to account for is the cost associated with brand damage. The costs associated with harming a reputation that we spent years and hundreds of thousands of dollars – if not millions – to craft. All can be lost in a single, poorly handled incident. And remember – these incidents are inevitable. You will go down.
Uptime monitoring solutions and application performance management tools are a form of preventive medicine. It’s like exercising. But when you get sick and you need some pharmaceutical grade medicines – that’s where something like Uptimely kicks in.
When the inevitable happens, and we have an application outage, it is crucial that we effectively communicate with our external customers to minimize damage to our brand and reputation. This isn’t the focus of uptime monitoring solutions and application performance management tools. These solutions are for alerting, and informing, internal IT staff. These tools are not meant to facilitate external communication with end-users. Yahoo did a very poor job of communicating with their end-users. This resulted to an uproar on Twitter, viral spread of negative word of mouth, and a major punch in the gut to Yahoo Mail’s reputation. The cost of such a poorly handled outage communication – priceless!
Yahoo is not unique in this regard. From our internal surveys, based on customers downloading our application status software, we know that less than 3 in 10 SaaS companies have a dedicated end user communication method outside of manually posting to twitter, sending out an e-mail, and posting on their website. Just researching this particular Yahoo Mail outage alone required me to jump between Twitter, Tumblr and Yahoo’s help page (their link is now dead) to try to figure out what happened. In fairness, there is a poorly placed status page but it doesn’t send SMS or e-mails (unless you go through Yahoo Mail, which was down – in other words don’t depend on your own systems to report downtime), it also doesn’t coordinate any customer discussion, overall uptime metrics or any type of historic service level reporting. Here is an example of how a status page should look.
It’s important to be proactive about communicating at the first sign of trouble. Often times IT or DevOps think they have the problem under control and they wait to convey issues, thinking it will be resolved quickly. Many well-known outages have a sudden spark and then the issue starts to spiral out of control, funneling its way through the system. Before you know it, you’re down for 24 hours, 1 day, 2 days, etc. If not handled properly from the start, suddenly what was an IT problem is now a PR nightmare. When you throw brand, loyalty, reputation and lost customers into the mix the potential losses are incalculable, and it could ultimately cost you your business.
