How Etsy reviews work, and why 4.9 stars does not mean what you think
A shop with 4.9 stars across three thousand reviews looks settled. Then you read the one star reviews and forty people describe the same broken thing. Both are true at once, and the reason is in how the average is built.
What the number counts
Every review counts the same. A buyer delighted with a two dollar sticker and a buyer sent the wrong item for eighty dollars each move the average by one review. Volume decides everything, so a shop with enough happy orders can carry a great many unhappy ones and still read 4.9.
Work it out and it stops being surprising. Three thousand reviews at 4.9 leaves room for roughly a hundred and fifty one star reviews. A hundred and fifty people is not a rounding error. It is a room full of them.
What the number leaves out
Most unhappy buyers never write anything. They ask for a refund, get it, and move on, and a resolved complaint leaves no mark on the rating at all. The reviews you can see are the ones by people annoyed enough to type.
Reviews also follow the listing rather than the shop's behavior. A seller who delists a product and relists it starts that product's reviews again. Nothing about the shop changed and the history is gone.
What to read instead
Sort to the one and two star reviews and read twenty of them. You are not looking for angry, you are looking for repeated. One person calling a necklace flimsy is one person. Nine people using the word flimsy about the same necklace is the necklace.
Then look only at reviews with photographs. Ten percent of buyers post one, and those ten percent settle in a second what a paragraph can only assert.
What we do with all this
We read every review on a shop, keep the ones that describe what arrived, and score the shop on how often buyers said they got something other than what was pictured. A star rating answers how many people were pleased. Ours answers a narrower question, which is whether the thing in the photograph is the thing in the box. How the score is built.