You need to see an accurate average Star Score for each company to make a smart choice. We optimize for accuracy better than any review site. ONE AllRatings super-composite Star Scores are optimized based on the attributes of each individual review, each company’s conduct, and each review site’s trustworthiness.

Customers Spent Their Money

We remove rating scores from reviewers who were not customers. Many review sources count these ratings, causing a distorted, inaccurate customer satisfaction star score. When you spend your money, you want what was promised. It’s misleading to count a 5-star score from a guy who loves the paint job on a roofing van he sees parked in a neighbor’s driveway the same as a 1-star score from that same neighbor who endured delays and shoddy work on her $10,000 roof project.

Each Customer’s Rating is Weighted by Recency and Analyzed for Legitimacy

Recent company star ratings are more predictive of near-term company performance. That’s why we weight each customer’s review score by recency. For example, reviews in the past year are weighted more than reviews from seven years ago. Individual reviews are analyzed for legitimacy to downweight questionable and remove obviously fake reviews.

Our Score Penalizes the Bad Actors

The conduct of a company matters. While most are mediocre to great, there are some really bad actors who cut corners and try to game rating systems. They may buy a bunch of reviews, or be unlicensed in a field that requires one, or have a horrendous record on handling complaints. We try to uncover these signals of bad conduct. We then downweight their rating score based on the severity of their conduct. It’s a smart way to keep you safe.

Review Sites Differ in Trustworthiness

We analyze the methodology and practices of each review site source for indicators of trustworthiness. How do they deal with fake reviews? Do they allow company owners to suppress real reviews? Are owners able to publicly respond to reviews? These and other data points create a source weighting and optimize the accuracy of ONE AllRatings scores.