Weighted ranking calculator (branches or products)
A branch with 3 responses averaging 5.0 shouldn’t beat one with 200 responses averaging 4.7. This ranking fixes that: it pulls rows with few responses toward the overall average. Paste one row per branch, product or employee with its name, average score and number of responses.
How many “borrowed” responses at the overall average each row gets. By default, the median number of responses.
The calculation runs in your browser. We don’t store what you type.
How it’s calculated
First the overall average C is calculated, weighted by responses: each row counts according to how many responses it has.
Each row’s adjusted score is (n × average + m × C) / (n + m), where n is its number of responses and m is the weight of the overall average. It’s like giving each row m “borrowed” responses scored at C.
By default, m is the median number of responses across your rows. You can change it: a larger m pulls scores closer to the overall average.
Rows are sorted by adjusted score. Rows with fewer than half of m responses are flagged “few responses”. You can separate columns with commas, tabs or semicolons, use decimal commas and paste a header row, which is skipped.
How to read the result
- A row with many responses barely changes: its adjusted score stays very close to its own average. A row with few responses moves a lot toward the overall average.
- The “Change” column shows how many positions each row moved up or down compared with a ranking sorted by the raw average alone.
- If a row flagged “few responses” ends up at the top or bottom, don’t reward or penalize it yet. Collect more responses first.
- The ranking sorts rows, but it doesn’t tell you whether the gaps are large. If two rows have very similar adjusted scores, treat them as tied.
Frequently asked questions
- What is a Bayesian average?
- It’s an average that starts from what you already know (the overall average) and moves toward each row’s own score as it gets more responses. That way, a handful of very high or very low scores isn’t enough to reach first or last place.
- What value of m should I use?
- The median number of responses works well in most cases. If you want to be stricter with rows that have few responses, raise it. If every row has plenty of responses, you can lower it.
- What does the “Change” column mean?
- How many positions each row moved compared with the ranking by simple average. A positive number means it moved up thanks to the adjustment, and a negative one means it moved down.
- How do I get the score and responses for each branch?
- With a satisfaction survey for each branch or product. In SurvIQ you can create one for free, share it with a QR code at each location and see the responses in charts in real time.
Know what you want to ask?
Build the survey in SurvIQ, share it by link, WhatsApp or QR code, and watch results come in live. The Free plan is free forever.
Create a free survey