Let’s say a friend of yours self-publishes a book on Amazon. She asks you to buy it, rate, maybe even gives you a copy. Naturally, she also hopes you’ll read it. Which you totally plan to do, as soon as you have a little time.

Meanwhile, though, she needs a few good reviews on Amazon just to get noticed. What would you do?

How about if you work at a restaurant, and your tips depend somewhat on how many people come in expecting good service. You have a couple regular customers that tell you how much they love the place. Would you ask them to post a review on Yelp?

Online reviews are becoming more and more part of everyday decision-making. Amazon, Netflix, and Bestbuy all provide a robust user-review platform which is very much part of why people use these services so extensively. Studies have shown a significant impact on restaurants from as little as a half-star difference in ratings on Yelp.

Like every form on online information, reviews are subject to abuse. People bash products they’ve never tried for reasons of politics or brand loyalty. A single bad experience can lead to an angry denunciation, that might be completely forgotten by the writer—but lives on online. Parties in a divorce will sometimes write fake negative reviews of their ex-partner’s lawyer because of the bitterness of the proceedings. Some studies have shown that the most negative reviews actually come from a companies biggest fans—the people who are often angry at change or disappointed in a specific product.

Businesses and professionals often assume that most negative reviews are either faked or wildly exaggerated, and sometimes respond when they shouldn’t. Google “Amy’s Baking Company” for a particularly bad example. This may lead some to try and game the system themselves. A hotel chain used a system that identified happy customers, via a survey, and gave them a link to a review site. Unhappy customers were not given the same link.

Further on down that slippery slope are “reputation management companies” that apparently plant fake positive reviews. Recently Edmunds.com sued one of these companies for placing over 2,200 fake reviews for 25 clients—automobile dealers—on its website.

Large review platforms spend a lot of time and effort trying to give more weight to honest user reviews. Over time, things tend to sort themselves out. This large-scale study noted that early reviews for hotels in a region tended to be negative, but over time as the population of reviews grew, a more accurate normal was achieved.

What is also interesting, but has not really been studied at this point, is the layer of active review-policing occurring on some platforms.  On Amazon, for instance, reviewers will often call out others they believe to be spammy or unduly negative. For the information seeker who wants to spend a little more time, there is a lot of information ABOUT the reviews, and the reviewers, as well.