Average Rating Calculator
Enter the number of reviews at each star level (1 through 5) and this calculator instantly computes the weighted average rating, the total review count, and the percentage share for each star tier. It also applies the Wilson score lower bound so you can see how your rating would hold up with more data. Results update as you type.
Formula
Worked example
A product has 120 five-star, 80 four-star, 30 three-star, 15 two-star, and 5 one-star reviews (250 total). Weighted sum = (5x120)+(4x80)+(3x30)+(2x15)+(1x5) = 600+320+90+30+5 = 1045. Average = 1045/250 = 4.18 stars. Wilson lower bound (treating 4+5 star as positive): (120+80)/250 = 0.80 positive rate, Wilson lower = 0.746, scaled to 5-star = 1+0.746x4 = 3.98.
How the weighted average rating is calculated
A simple average adds all individual ratings and divides by the count, which gives you the same result whether someone gave 1 star or 5 stars. A weighted average rating goes further: it multiplies each star level by the number of votes it received, sums those products, then divides by the total number of votes. This is sometimes called the weighted mean. For a 5-star scale, the formula is: Average Rating = (5 x votes5 + 4 x votes4 + 3 x votes3 + 2 x votes2 + 1 x votes1) divided by the total vote count. Because the weights (star values) and the frequencies (vote counts) are both used, a single 5-star review has exactly five times the impact on the numerator that a single 1-star review does, but the denominator grows by one for each regardless.
Why the raw average is not enough for ranking
Two products can share the same raw average yet be very different in quality signal. A product with 4 reviews averaging 4.5 is far less reliable than one with 400 reviews averaging 4.3. The Wilson score lower bound addresses this by estimating the lowest plausible true rating at 95% confidence given the observed data. It shrinks toward 50% when the sample is small and converges on the true rate as volume grows. Amazon and Reddit both use variants of this approach to surface items that are genuinely good rather than items that just got lucky with a few early reviews. This calculator reports the Wilson lower bound so you can use it as a conservative ranking metric when comparing products or services with very different review volumes.
What the per-star percentages reveal
The percentage breakdown tells a richer story than the average alone. A J-shaped distribution, many 5-star and many 1-star reviews with few in between, is typical of polarizing products and often signals that quality is inconsistent or that the product appeals strongly to a specific niche while disappointing others. A bell-shaped distribution clustered around 3-4 stars suggests a product that is broadly acceptable but lacks standout qualities. A distribution dominated by 4- and 5-star reviews with few 1-star entries is the profile most associated with consistent quality. Looking at the star breakdown helps you identify whether a negative average is driven by a systematic problem or by a vocal minority.
Industry benchmarks and what counts as a good rating
A rating of 4.0 or higher on a 5-star scale is generally the floor for products and services that convert well. On Google and Amazon, the 4.2 to 4.5 range has been shown in multiple studies to be the sweet spot where consumer trust peaks, because ratings above 4.7 or 4.8 can actually trigger skepticism about authenticity. On 10-point scales (common in hotel and software review sites), the equivalent competitive range is roughly 8.0 to 9.2. Below those floors, conversion rates drop sharply. The reference table above shows typical ranges and thresholds across major review categories. Note that a lower Wilson score at high volume is often more trustworthy than a higher raw average at low volume.
Industry average star ratings by sector
| Industry / Platform | Typical range | Competitive threshold |
|---|---|---|
| E-commerce products (Amazon) | 3.8 - 4.5 | 4.0+ |
| Restaurants (Yelp, Google) | 3.5 - 4.5 | 4.2+ |
| Mobile apps (App Store, Google Play) | 3.7 - 4.6 | 4.0+ |
| Hotels (TripAdvisor, Booking.com) | 7.5 - 9.0 | 8.0+ (10-point) |
| SaaS / software (G2, Capterra) | 3.8 - 4.7 | 4.2+ |
| Local services (Google Maps) | 3.9 - 4.8 | 4.3+ |
Typical weighted average ratings observed across major review platforms. Ratings above 4.0 are considered competitive in most categories.
Frequently asked questions
What is the formula for average star rating?
Average rating = (sum of each star level multiplied by its vote count) divided by the total number of votes. For a 5-star system with vote counts r1 through r5, the formula is: (5r5 + 4r4 + 3r3 + 2r2 + r1) / (r5 + r4 + r3 + r2 + r1). This is a weighted mean where each star value is weighted by how many reviewers chose it.
What is the Wilson score lower bound and why does it matter?
The Wilson score lower bound is a statistically rigorous estimate of the lowest plausible true rating at a given confidence level (this calculator uses 95%). It penalizes averages based on small samples, so a product with 5 reviews averaging 4.8 will have a much lower Wilson score than one with 500 reviews averaging 4.6. Many large platforms use it (or similar Bayesian approaches) to rank listings so that items with a track record of good ratings surface ahead of lucky newcomers.
How many reviews do I need for my average to be reliable?
There is no universal threshold, but 30 reviews is a commonly cited floor for a stable signal. Below 10 reviews, a single extreme rating can shift the average by half a star or more. Above 100 reviews, the weighted average is generally stable and the Wilson score converges close to the raw average. For high-stakes decisions such as purchasing or ranking supplier bids, aim for at least 50 verified reviews before treating the average as representative.
Is a 4.5 rating better than a 5.0 rating?
Sometimes, yes. Research from multiple e-commerce platforms shows that consumer trust peaks in the 4.2 to 4.5 range rather than at a perfect 5.0. A flawless score with few reviews can look suspicious to buyers who assume all negative reviews were removed. A well-established 4.3 or 4.4 average with hundreds of reviews often converts better than a pristine 5.0 with only a handful.
Can I use this calculator for a 10-point scale?
Yes. Switch the "Rating scale" input to "10-point (1 to 10)" and enter the vote counts for each level from 1 through 10. The weighted average, total votes, and Wilson score lower bound (scaled to 10) are all computed for you. This is useful for hotel, software, and NPS-adjacent scales that run from 1 to 10.
What does the J-shaped distribution mean for my product?
A J-shaped (or bimodal) distribution is one where most reviews are either very high (5-star) or very low (1-star), with fewer in the middle. It is common with polarizing products: one audience loves it while another is disappointed, or quality control is inconsistent. If you see it, read the 1-star and 5-star reviews separately to understand who is unhappy and why, rather than focusing on the average alone.