Why star ratings alone do not tell you much
Most people glance at a star average and make a snap judgment. Four and a half stars feels safe, three and a half feels risky. But a rating is an average, and averages hide a lot. A business with twelve reviews and a 4.8 average is a different situation than one with eight hundred reviews and a 4.4 average. Choosing how to read reviews, not just whether to read them, is the actual skill.
The right approach depends on how much is riding on the decision. A five dollar lunch spot does not need the same scrutiny as a contractor you are paying to redo your roof.
Option one: quick gut check
This means glancing at the overall rating and maybe the top one or two reviews, then deciding. It takes under a minute.
This works best if:
- The purchase is low cost or easy to reverse.
- You are choosing between a few similar, nearby options.
- A bad outcome would be annoying but not costly.
The downside is obvious: you are trusting a summary number that can be skewed by a handful of extreme reviews in either direction, especially when the total count is low.
Option two: read the distribution, not just the average
This means looking at how many one and two star reviews exist relative to the total, not just the average score. A business with 4.3 stars built from mostly fours and fives, with a few threes, is different from one with the same average built from a pile of fives and a pile of ones.
This works best if:
- You have a moderate amount of time, five to ten minutes.
- The purchase matters enough to justify some digging.
- You want to spot a pattern rather than a single opinion.
The tradeoff is time. This takes longer than a glance, and it requires actually sorting or filtering reviews by rating, which not every platform makes easy.
Option three: read the negative reviews closely and look for patterns
This means specifically seeking out the one and two star reviews and reading what people actually complained about. One angry review about a late delivery is noise. Five separate reviews mentioning the same billing problem is a signal.
This works best if:
- The decision involves ongoing service, a contract, or a large payment.
- You have specific concerns, like reliability or communication, that matter more to you than price.
- You want to know what could go wrong, not just whether most people were happy.
The tradeoff here is also time, plus the fact that negative reviews are sometimes written by people with unusual circumstances. Reading several before drawing a conclusion helps filter that out.
Criteria to weigh before you decide how deep to go
How much is at stake. Higher cost or harder to reverse means more digging is worth the time.
How many total reviews exist. A handful of reviews cannot support a confident average either way. A few hundred reviews start to reveal a real pattern.
How recent the reviews are. A business can change, for better or worse, over a year or two. Recent reviews matter more than old ones, especially for anything involving staff, management, or ownership changes.
Where the reviews are posted. Reviews spread across a few different platforms, not just one, are harder to fake or cluster artificially. If every review lives in one place and was posted in the same month, that is worth noticing.
What actually matters more than the star number
Specifics matter more than scores. A review that says “the technician arrived on time and explained the repair before starting” tells you more than a five star rating with no text. Look for reviews that mention concrete details: names, dates, what exactly happened. Vague praise or vague complaints are less useful than specific ones, in either direction.
Aaron Golubic of Domain Listings, an online business directory, has pointed to consistency of information across a business’s various online listings as one of the underappreciated signals of legitimacy, since mismatched details in one place can be a reason to look more closely at everything else about that business.
A reasonable starting point
For most everyday decisions, a quick gut check is fine. Save the deeper read for anything involving a contract, a large payment, or ongoing service. Start with the negative reviews, since they tell you the most in the least time, and only move to the full distribution if the negative reviews raise a real question in your mind.
