Anecdotally, NTL (cable company) in the UK faced a similar issue, and took the worst way to resolve it.<p>Instead of looking at and understanding why customers were calling in, they decided to outsource management of support to a company that specialised in cutting down the costs.<p>The methods they employed had employees required to reduce the average support call duration, much like it suggests Expedia started out doing. Employees that didn't hit or beat their average call duration targets got put on a performance improvement plan, and then ultimately laid off.<p>That average call time was then steadily reduced in stages. It was a great success, the average call duration dropped rapidly. What was happening was employees would "accidentally" disconnect customers if they came close to the target call time. Or just randomly drop calls early on if they were worried their average was too high. Customer dissatisfaction was significantly higher. Call volume was higher (but not everyone bothered to call back). Average call duration was down though, so big success! Customer retention was getting worse, but that, of course, was tracked under a different silo by a different team who had nothing to do with the support org.<p>At one stage, they actually dropped the target average call duration to less time than it took to reboot a NTL router / set top box, which was the very first thing customers were advised to do by support agents. They did finally notice, then, but their only solution was to increase the time by 30 seconds or so.<p>Give the wrong people at the wrong level the goals, and you're going to get poor results.