50% Discount for New Users! Copy code Here
cross

Penny for Your Thoughts: The Dangers of Review Based Decision Making

In this article you will learn about:

  1. The shocking truth behind 5-star reviews: Are you being fooled by "review farms"?
  2. Pay-to-play: How review sites are rigging the game against you
  3. Why trusting online reviews for B2B partners could sink your business

Introduction

Have you ever thought to yourself: “how is it possible that over 10,000 people have rated this (insert product or service) five stars? How in the heck am I supposed to make a decision with this?” Us too. Let’s explore the problems with customer reviews, their impact, and what you can do to make better decisions, especially when it comes to choosing critical B2B professional services partners. 

As a society, we have historically relied on the opinion of others in all facets of our lives. From who we pick as romantic partners, to what kind of toaster we should choose. And, as we choose more and more of these online, there has been a deluge of data on the opinions of others in all of these decisions. In the earlier days of the internet, these reviews could be helpful; people who had purchased that toaster or that blender wrote reviews and gave their opinions on what worked and what didn’t. Today, the game has changed. What was once a reliable and helpful transfer of trust has become a manipulative tool for marketing. 

What’s wrong with reviews? 

To illustrate this point, an amazon.com search for a toaster yields the following top 3 results: 

Figure 1: manipulated search results

The key question is: are these ratings helping anyone anymore? Out of these three, two are “sponsored” - conflating the seller’s paid results with the best options, and all three have over 15,000 reviews with an average of 4+ stars - does this smell fishy to anybody else? Did fifteen-thousand actual humans care enough about a $16 purchase to write a real and thoughtful review? Doubtful. This is indicative of some of the key problems we’ll explore in this post: 

  1. The reliability of reviews in the age of review farms
  2. Pay for promotion and its impact on decision making
  3. Selection bias in reviews

The impact on professional services vendor selection 

Taking this problem into our area of expertise, professional services (proserv), let’s examine these issues and their impact on how businesses find proserv partners:

  1. The rise of review farms: in recent years, to meet the need of businesses to appear well-liked online, a new farming service has emerged: click, bot, review, and content farms. Companies where businesses can hire organizations in emerging economies to create false reviews, clicks, likes, etc. using humans, bots, or both (see figure 2 below for what these farms can look like). These farms have been used to manipulative ends in all industries, from toaster reviews, to sites such as G2, TrustPilot, and more, where companies in industries from software to services pay these farms to create falsified reviews to boost their reputation. One good explanation of how these bot farms work can be found in THIS TikTok post from a former bot farm operator.
  2. Pay for promotion: the same problem illustrated by the “sponsored” toasters above is rampant in the professional services industry. Paying for plans on sites like G2, Trustpilot, and Clutch gives you greater visibility among potential buyers. While this practice of pay for promotion is nearly as old as time, it also goes against the root purpose of these sites, which is to help people with unbiased reviews of potential software and services. It also works against the newcomers and smaller players, who may provide exactly what you’re searching for, but don’t yet have the capital to purchase the visibility or quantity of reviews that will get them noticed by potential customers.
  3. Pay for reviews: Review sites run on…reviews. How you get those reviews makes all the difference. Aside from simply paying for click / bot farms, sites like this and their clients also rely on paying users to provide feedback. There are many issues with this, chief among them that the user then feels an obligation to the provider to write a positive review, since they are being paid (see figure 3 below), and more importantly, that this can be easily manipulated by the service provider and the review site by asking for reviews primarily of satisfied users rather than those who have complained, left the site, etc. While these sites list robust content moderation guidance, which should adhere to policy such as FTC guidance on not segmenting reviews, the reality is that the data speaks for itself. The distribution of reviews skews very heavily towards 4+ stars out of 5.

Figure 2: review farm in action

Figure 3: Pay for reviews 

How can we solve it?

This is where we believe that AI can serve a valuable role as an unbiased aggregator and analyzer of all the public data points on a given company. We built sc0red to contend with just this challenge. Similar to gold standard rating guides such as Consumer Reports, the Wirecutter, the Michelin Guide, etc., sc0red uses rating criteria from industry veterans, scaled with the power of AI, to provide helpful, clear, and evidence based ratings. sc0red works by: 

  1. Identifying the kind of assessment questions that should be asked about different types of providers, from experts in that field. For instance, the 30-40 key questions that should be answered about software engineering firms
  2. Using our advanced AI to scan the internet for over 1,000 data points to answer each of these questions, whether the answer is good, bad, or unclear 
  3. Displaying this information as unbiased and simple answers, with sources for each question 
  4. Never asking service providers pay for increased visibility on the site
  5. Only increasing a service provider’s rating based on verification of their capabilities from submitted and examined proof

Let’s take an example to compare and contrast a review site with sc0red. Looking at the top mobile app development companies on G2.com, we see the following: 

Figure 4: top mobile app providers from G2.com 

Taking one of the top providers we see a perfect 5 star score based on the reviews provided. When asked “what do you dislike about the company”, the following is common:

  1. “There is not a single point that I can suggest to improve, they have already top-class services provided”
  2. “I really don't have any dislikes about them”
  3. “Based on my experience, there is not a single negative point about their work”

While nice, it doesn’t exactly help the service seeker make an informed decision. So what does sc0red have to say? 

Figure 5: sc0red review of a top G2 pick

While not bad at 3 out of 5 stars, it highlights several areas of assessment that the service seeker may want to delve deeper into. This gives them insight into potential blind spots that could be missed if only going off biased reviews. 

Conclusion

We are in an epidemic of fictitious and misleading review data, forcing service seekers to find alternative means of finding and selecting partners for critical projects. Especially when the decision is more important than finding the right toaster, but rather - will this partner enable my business to grow or threaten its viability. We need a better way to sort the wheat from the chaff, the sheep from the goats, or the raisins from the bran. Today’s review sites have become an “advocate for hire” rather than an unbiased guide to find your best-fit proserv partner, and we need a better way. 

We hope that you give sc0red a try and drive progress through partnerships.

Sources

  1. Everything you need to know about click farms. Cheq
  2. Comparing Software Review Sites, Trustloop
  3. Do product review sites like G2 really matter? Medium
  4. G2 content moderation guidelines
  5. Featuring Online Customer Reviews: A Guide for Platforms. Federal Trade Commission (FTC)