Shills on travel Web sites

I’m quoted in the BBC’s online magazine about the problems of shills (impostors associated with a business who pretend to be ordinary customers while talking up the company and/or talking down its competitors) on travel Web sites, a phenomenon I wrote about most recently here a few months ago. The comments on the BBC Web site are worth reading for some examples of readers’ experiences with shills.

As usual, there was only room in the published story for a few snippets from a longer interview. Here’s more:

BBC: What is the damage done by shilling and how prevalent is it?

EH: Obviously, the damage to travelers is to those who rely uncritically on anonymous reviews. It can also reward disreputable travel companies (hotels, etc.), and damage the reputations of good ones. Knowledge of shilling also damages the business of review Web sites, who therefore have an interest in reassuring the public that they have eliminated shilling. I think they are sincere in wanting to eliminate shilling, but their busienss imperative is to persuade the public thay have done so, lest people start looking elsewhere (professionally written guidebooks, or recommendations from a trusted travel agent who has actually inspected the hotel or spoken to recent bona fide guests) for information and advice.

BBC: What can sites do to stop shilling?

EH: The only way — easy but very, very, expensive — to prevent shilling is to publish reviews by paid, known, staff, vetted by editors, rather than “user-generated” content.

Reviews by unknown people can be useful, but the risk of shilling is inherent in user-generated content. Claims by Web site publishers that they can detect, prevent, or remove all or most shill reviews are nonsense. Publishers whose business models for their Web sites depend on not paying professional writers or reviewers have to claim, “Trust us”, but you shouldn’t believe them. Sure, some clumsy shills are obvious, but many aren’t.

The question isn’t whether shilling can be prevented, but how to use collections of information that you know include some shills mixed in with the legitimate amateur reviews, and how to balance your use of user reviews and reviews by professionals (e.g. in guidebooks). I use both myself.

BBC: What can users of the sites do to spot shills?

What can you do? Don’t believe everything you read. Shills are most prevalent on accommodations review sites, where a few glowing reviews by shills (or negative reviews of the competion) can make the difference between profit and loss for a small hotel, guesthouse, or bed & breakfast. If there are only a few reviews of a place, assume that there is a good chance they’ve been planted by friends or foes. If there are many reviews, act like a statistician, and start by dropping the outliers. And always remember that there’s no accounting for taste: Even impartial reviewers may have a completely different reaction than you to the same conditions or situation. Every review reflects the reviewer as well the thing reviewed — and on the Web (except perhaps in the blog of someone you know personally), you only know how reviewers describe themselves, not what you would think of them if you met.

You can’t count on spotting shill reviews. You have to assume that anonymous reviews, or any content not explicitly labeled as being from a named, identifiable, trusted source (e.g. a bylined review credited to a reputable news or guidebook publisher), is or might be advertising.

Keep in mind, of course, that both favorable and unfavorable reviews (directed at competitors) can be planted.

Advertising brochures are useful, as are anonymous reviews, but I always try to balance them with at least some content from a known and independent source. That’s really the most important thing.

For what it’s worth, I think that Web sites that mix advertisements and editorial content (without the clear distinctions traditional in, say, newspapers), or that fail to identify the source of reviews or descriptions (are they by independent reviewers, or provided by the businesses being reviewed?) are as much of a problem as shills.

Another problem as serious as shilling is user reviews (or sometimes guidebook reviews, especially in guidebooks written by commitee or from publishers who edit out the personal voice of the writer/reviewer) that draw conclusions without giving you enough background and perspective. You may love a place for the same attributes that led someone else to hate it, or vice versa. Withought knowing the speaker and their tastes, it’s very hard to know what weight to give to their advice, or how to interpret it. I prefer guidebooks written by individual writers with identifiable personalities and explicitly stated points of view.

[Update: More from the Associated Press: “Arthur Frommer, founder of the Frommer’s travel guides, said travelers should rely on the advice of experts, such as guidebook writers or journalists. Frommer said he had to discontinue a popular feature in some of his guidebooks that included readers’ selections, largely because he could not be certain they did not come from businesses. ‘Find write-ups by professionals whose judgments you trust and rely on that,” Frommer said. “I would never rely on the judgment of amateurs.’”]