PERCENT_RANK & NTILE: Segmenting Your Data

Welcome to **Day 68**. Today we learn how to answer the question: "Which bucket does this row fall into?"

Percentiles: PERCENT_RANK()

`PERCENT_RANK` returns a number between 0 and 1 representing where a row sits as a percentage.

0.99 = Top 1%.

0.50 = Median.

SELECT

name,

salary,

PERCENT_RANK() OVER(ORDER BY salary) as salary_percentile

FROM employees;

`NTILE(n)` slices your data into `n` equal groups. This is perfect for the marketing team who wants to send different emails to:

**Tier 1 (Top 25%)**: Big Spender discount.

**Tier 4 (Bottom 25%)**: "We miss you" discount.

SELECT

customer_id,

NTILE(4) OVER(ORDER BY total_spent DESC) as customer_tier

FROM customers;

If you have 100 students and you want to give A's to the top 10, use `NTILE(10)`. The first bucket contains your straight-A students!

Segment your product catalog into 10 price "Deciles" using `NTILE(10)`.

*Day 69: FIRST_VALUE & LAST_VALUE—Retrieving endpoints.*