When working with data in Tableau Prep, you might want to divide a continuous measure into groups of the same size for analysis (quartiles, deciles etc.) This is where tiles come in.
Tiles allow you to divide a dataset into a specified number of evenly distributed groups based on a numeric field. Each group contains (roughly) the same number of records, making tiles a useful way to segment data for analysis.
What Are Tiles?
A tile assigns each row in your dataset to a group based on its position when the data is sorted by a chosen field.
For example, if you create 4 tiles on a dataset sorted by sales, each tile will contain approximately 25% of the records:
- Tile 1 → lowest values
- Tile 2 → lower-middle
- Tile 3 → upper-middle
- Tile 4 → highest values
This is similar to ranking, but instead of assigning a unique rank to each row, tiles group rows into buckets.
When Would You Use Tiles?
Tiles are particularly useful when you want to analyse relative performance or distribution. Common use cases include:
- Quartiles (4 tiles): splitting data into four equal groups
- Deciles (10 tiles): dividing data into ten groups for more granular analysis
- Percentile-style groupings: understanding where values fall within a distribution
- Customer segmentation: grouping customers by spend, engagement, or frequency
Because each tile contains a similar number of records, this approach is often more balanced than using fixed bins.
Tiles vs Bins
It’s worth distinguishing tiles from bins, as they serve different purposes:
- Tiles: equal number of records per group
- Bins: equal range of values per group
For example:
- Tiles ensure each group has similar counts
- Bins ensure each group covers the same value range
Which you use depends on whether you care more about distribution or value intervals.
How Tiles Work in Tableau Prep
In Tableau Prep, tiles are created within a clean step. You choose:
- the numeric field to base the tiles on
- the number of tiles you want to create
Prep then sorts the data and assigns each row to a tile automatically.
In the video below, I’ll walk through exactly how to set this up and show how tiles can be used to quickly segment a dataset.