A report released in 2013 by Emerson Network Power revealed that just one minute of unplanned system downtime costs the average company $5,000. And that was last year; applications are only getting more complex and harder to handle. Conventionally, a runbook would provide detailed instructions for what to do when something breaks, but modern applications now operate on too large a scale to anticipate every problem. Technically, none of the companies in this slide show did anything wrong; they just couldn’t plan for the causes that landed them on this list. All have global, Web-scale architectures, making it impossible for development teams to act before frustrated users speak up on social media. Is there a solution to this daunting task? There is. Artificial intelligence-based monitoring tracks down problems before they result in outages, giving people advanced notice of potential issues and providing peace of mind for an increasingly complex landscape. This slide show uses eWEEK reporting and industry information and commentary from Alois Reitbauer, chief evangelist for artificial intelligence-based performance monitoring and analytics provider Ruxit, to break down the top 10 avoidable application outages of 2014.