This tool generates synthetic survey response datasets for testing analytics and visualisation tools. Each row carries a Likert satisfaction score, a single-choice recommendation answer, and a short free-text comment, and the whole set exports as a clean CSV. It lets you exercise charts, pivots, and sentiment views with realistic-looking data without using any real respondents.
How it works
For each requested row the generator builds four fields:
- A respondent id in the form
R0001,R0002, and so on. - A satisfaction rating from 1 to 5, sampled from a weighted distribution that skews slightly positive so the spread resembles a real survey rather than a flat random scatter.
- A single-choice recommendation answer drawn from a fixed five-point scale.
- A free-text comment assembled from an opener, an adjective, and a closer fragment.
The dataset is rendered as a preview table and can be copied as CSV. Every text field is escaped per RFC 4180 — fields containing a comma, quote, or newline are wrapped in double quotes and any internal quotes are doubled — so the output imports without corrupting columns.
Why the distribution is weighted
A completely flat uniform distribution (equal probability of 1, 2, 3, 4, or 5) would produce a mean of exactly 3.0 and look suspiciously balanced. Real satisfaction surveys typically skew slightly toward higher ratings, with more 4s and 5s than 1s and 2s, and a mean closer to 3.5 to 4.0. The weighted distribution here mirrors that pattern so your visualisation receives data that looks like it came from real users rather than a dice roll.
Column reference
| Column | Content |
|---|---|
respondent_id | Unique ID: R0001, R0002, … |
satisfaction | Integer 1 to 5 (weighted, skews positive) |
recommendation | Single-choice text from a 5-point scale |
comment | Short constructed sentence |
The comment text is assembled from fragment banks that are loosely correlated with the satisfaction score — higher scores tend to produce more positive openers and adjectives. This makes the text look internally consistent when someone reads the CSV row by row, rather than obviously random.
When to use this versus the NPS generator
Both tools produce synthetic survey data, but they serve different formats:
- This tool generates Likert-scale satisfaction surveys with a 1–5 rating — useful for customer satisfaction (CSAT) dashboards, product-feedback tools, and heat-map visualisations.
- The NPS generator produces Net Promoter Score datasets with 0–10 scores mapped to Promoter/Passive/Detractor — useful for NPS dashboard components that show the NPS formula result.
If your analytics tool ingests both types, use both generators to verify each chart component separately.
Tips and notes
- The mean satisfaction shown above the table is a quick sanity check that your charts read the rating column correctly — it should be around 3.5 to 4.0, not exactly 3.
- Generate a few hundred rows when you need realistic distributions for histograms or box plots.
- The preview shows the first 25 rows for speed; the Copy CSV button always exports the full set.
- Everything runs locally in your browser, so regenerate as often as you like with no network calls.