Deepak Desai, CEO of GlobalEnglish, a provider of online language training tools to corporate customers, pays close attention to the feedback he gets from users. And two years ago, it became clear that the users had spoken: They were commenting frequently, online and in person, that the training required too much repetition of specific phrases and vocabulary. Desai and his team pored over the comments, analyzed the problem, and took decisive action that customers applauded: They didn't change a thing.
Today, everyone wants to create the customer-driven company, and that means getting lots of customer feedback. Hey, no problem. The Internet provides a fire hose of customer data: website clicks, online product evaluations, surveys, chat rooms, e-mail, instant messages, even Skype calls and video chats. You can hear more from your customers than you do from your family. The advantages of all this feedback are obvious. Facebook, for example, learned quickly how people felt after it offered a feature that let members monitor changes on other people's pages. Users hated it.
But often overlooked is the fact that all this readily gleaned customer feedback can lead you astray. Take GlobalEnglish. It's not that Desai doesn't act on customer comments and complaints. When users asked for more help writing business documents in English, for example, the company created an online writing center with customizable sample documents. But Desai casts a skeptical eye over the feedback before doing anything. In the case of users complaining about too much repetition, Desai recognized that they were expressing natural frustration with the difficult and often monotonous drills essential to mastering a new language. "You have to be careful to not take all feedback seriously," says Desai. "My son complains about his homework, but I wouldn't be doing him a favor by excusing him from it." And sure enough, customer satisfaction with GlobalEnglish is high precisely because users eventually realize all those language drills yield results.
Marketers have long known that sometimes customers hate you for listening to them. Think New Coke, which tested off the charts with soda drinkers up until the moment it was rolled out, when it flopped spectacularly. The problem crops up even with meticulously designed focus groups, and things only get dicier when you're dealing with the raw, random river of Internet feedback. For starters, you can end up listening to the wrong customers. Online, the loudest and most-oft-heard customers might think very differently from the heart of your customer base. Sometimes a squeaky wheel is simply misaligned and heading in the wrong direction. If Microsoft (NASDAQ:MSFT) took its cues from the constant stream of disdain that it attracts from the cranky, iconoclastic message posters who tend to dominate online chatter, the company might have killed its Windows software long ago--though Windows has, of course, utterly dominated mainstream corporate and home computing.
Taking customer feedback with a grain of salt may sound easy enough, but entrepreneurs who aren't used to navigating a blizzard of comments are often quick to overreact, notes Gary Rhoads, a business professor at Brigham Young University who studies customer satisfaction and who co-founded and now consults to a company called Allegiance, which provides online tools for managing feedback. "A manager will see a comment and say, 'Oh, my gosh, did you see what this customer said?" Rhoads says. "But a lot of the time it's just not representative of the customer population."
That problem emerged recently at Amarr, a garage door manufacturer in Winston-Salem, North Carolina. Amarr gets plenty of e-mail from contractors, but Steve Crawford, the company's IT chief, sees it as an unreliable gauge of customer opinion. "The people who buy and install our doors are mostly guys with pickup trucks, and half of them don't even have e-mail," says Crawford. "Even if they have it, their careers hinge on being able to get a door installed in two hours. They're not going to be sending off any e-mail notes about it." To get more representative feedback, the company instead relies on face time with customers in the field and in organized meetings.
Queensboro, an Internet-based apparel vendor in Wilmington, North Carolina, wrestles with similar questions about its feedback. Founder and CEO Fred Meyers notes that in addition to collecting e-mail notes and comments posted on the company's website, Queensboro also e-mails surveys to customers two days after an order arrives. About 35 percent respond. Meyers and other top managers meet weekly to sift through this cornucopia of comments and are sometimes confronted by quality complaints about the company's least expensive shirts--custom-logoed polos that sell for as little as $7.95 and indeed are inferior to other items. Queensboro wants to be perceived as a high-quality vendor, and its first reaction was to be worried; some managers even argued that the company should discontinue the cheap polos. But then they realized: The shirts cost eight bucks. Buyers who expect premium quality at that price probably aren't good candidates to become bigger-spending customers who buy higher-quality shirts. "We have to evaluate these sorts of comments in the context of what the customer is buying," says Meyers.
Online feedback is a good way for a company to get in on "crowd-sourcing"--that is, letting the customer base help drive the direction and design of new products and services. There's just one catch: If the crowd seizes on an idea it loves, it will pressure you to adopt it, even if the idea is costly and at odds with your own strategic thinking. After all, the crowd may know what it wants, but it isn't running a cost-benefit analysis on your behalf. Consider Jellyfish, a Middleton, Wisconsin, social shopping website that runs a daily hourlong game-show-like session for bargain hunters. Jellyfish employees mix into the crowd to solicit suggestions. The most popular: Run the session 24-7. "We've been hit with hundreds of comments asking for it," says Jellyfish CEO Brian Wiegand. "But it's a radical request that requires building in more support, inventory, and scheduling complexity." So far, the company has held off on committing to the idea.
Indeed, one danger of focusing on what customers ask for in feedback is that a company might stop trying to come up with bright new ideas that its customers aren't asking for. Apple's (NASDAQ:AAPL) customers weren't inundating it with requests for a hand-held music player when it came up with the iPod. "Companies that are overly customer-reactive can fail to think about the big picture," says Brigham Young's Rhoads. "You can't ignore what customers are saying, but you don't want to kill innovation by only responding to customers."
It would be nuts not to take advantage of the ability to gather as much customer feedback in a day as a company might have gotten in a year a decade ago. But good luck figuring out which feedback to act on, and which actions to take. In February, Dell (NASDAQ:DELL) put up a site to let the public post and vote on ideas for Dell computers. The company will adopt the popular suggestion to offer the Linux operating system, though it's been costly to make that happen. Dell isn't moving so fast on another customer request: PCs with cupholders.
Contributing editor David H. Freedman is a Boston-based author of several books about business and technology.