Garbage or Gold?

A recent article in the Sunday Times ofEngland points out the dark underbelly of user-contributed content. An undercover reporter posing as a salesperson for a marketing company called a number of hotels and offered to post favorable reviews about them on user-driven travel websites in exchange for a fee. None of the hotels contacted dismissed the proposition out of hand, and many of them were more than a little intrigued.

That's just one unseemly aspect of what the Sunday Times sees as a larger problem: user-contributed content in the travel industry is becoming more and more widespread and popular, forcing companies that conduct their own independent ratings to either adopt user-generated ratings or go out of business. This might be dismissed as the invisible hand of the marketplace at work, except for one major issue: in many cases, bad data is replacing good data.

The Times article found a number of examples of hotel owners submitting reviews on their own properties, and notes that for most travel sites even two or three reviews can radically alter the rating for a hotel for better or worse, creating fertile playground for fraudsters. In an even murkier twist, some hotels are now offering various incentives to guests who post positive online reviews. The newspaper also contacted a number of travel website operators that claimed to monitor all user-supplied content and found that some weren’t living up to that promise, Those that did monitor all user-submitted content admitted they couldn’t really spot bogus submissions. There are even cases now of hotels threatening legal action against their guests who chose to post legitimate, negative reviews.

What's going on here is that these travel sites are beginning to have real influence in the marketplace, with significant financial consequences for hotels. Hotel owners are frustrated by a system they don't consider fair, where just one or two dissatisfied guests can totally trash a hotel's reputation. Worse yet, the system is wide open to many forms of fraud and abuse. Operators of these websites are publishing information they can't vet, and consumers continue to lap up all free and easily accessible content with no apparent concern about its provenance or accuracy.

With user-generated content and community now two of the hottest concepts in the data publishing business, it's important that we all take a look at the critical issue of quality, because the current "garbage in, garbage out" model can't be a basis for long-term success.