Creating Custom Metrics in Apache Superset

  

Modern data analysis is not just about viewing raw tables or basic aggregations, it’s about defining business-specific metrics that reflect how an organization operates. This is where custom metrics in Apache Superset become extremely powerful.

 

Out of the box, Superset provides standard aggregations like SUM, COUNT, and AVG. While useful, they often fall short when you need deeper insights such as customer behavior trends, operational efficiency, or performance KPIs that are unique to your domain.

 

Custom metrics allow you to go beyond these limitations by writing SQL-based expressions directly at the dataset level. Once defined, these metrics become reusable building blocks that can be used across multiple charts and dashboards without rewriting logic every time.

 

In this topic, we will explore how to create and use custom metrics in Apache Superset using a practical dataset. You will learn how simple SQL expressions can be transformed into meaningful business indicators, enabling more insightful and decision-ready dashboards.

 

To demonstrate the examples, let’s use following customer orders dataset.

 

customer_orders.csv

order_id,customer_id,order_date,delivery_date,order_status,order_value
1,C001,2026-01-01,2026-01-03,Delivered,1200
2,C002,2026-01-02,2026-01-06,Delivered,800
3,C001,2026-01-03,2026-01-06,Delivered,1500
4,C003,2026-01-04,,Cancelled,600
5,C002,2026-01-05,2026-01-09,Delivered,950
6,C004,2026-01-06,2026-01-08,Delivered,400
7,C001,2026-01-07,2026-01-10,Delivered,700
8,C003,2026-01-08,2026-01-12,Delivered,1100
9,C005,2026-01-09,,Cancelled,500
10,C006,2026-01-10,2026-01-13,Delivered,1300
11,C002,2026-01-11,2026-01-15,Delivered,900
12,C007,2026-01-12,2026-01-14,Delivered,450
13,C001,2026-01-13,2026-01-17,Delivered,1600
14,C003,2026-01-14,2026-01-18,Delivered,700
15,C008,2026-01-15,,Cancelled,650
16,C004,2026-01-16,2026-01-18,Delivered,300
17,C002,2026-01-17,2026-01-21,Delivered,1050
18,C005,2026-01-18,2026-01-20,Delivered,980
19,C001,2026-01-19,2026-01-22,Delivered,1250
20,C009,2026-01-20,,Cancelled,550
21,C006,2026-01-21,2026-01-25,Delivered,1400
22,C003,2026-01-22,2026-01-26,Delivered,820
23,C002,2026-01-23,2026-01-27,Delivered,930
24,C007,2026-01-24,2026-01-26,Delivered,470
25,C001,2026-01-25,2026-01-28,Delivered,1700
26,C004,2026-01-26,,Cancelled,600
27,C005,2026-01-27,2026-01-30,Delivered,990
28,C008,2026-01-28,2026-01-31,Delivered,750
29,C006,2026-01-29,2026-02-02,Delivered,1350
30,C009,2026-01-30,2026-02-03,Delivered,500
31,C002,2026-01-31,,Cancelled,880
32,C001,2026-02-01,2026-02-04,Delivered,1180
33,C003,2026-02-02,2026-02-06,Delivered,720
34,C004,2026-02-03,2026-02-05,Delivered,410
35,C005,2026-02-04,2026-02-08,Delivered,970
36,C006,2026-02-05,,Cancelled,1450
37,C007,2026-02-06,2026-02-09,Delivered,460
38,C008,2026-02-07,2026-02-10,Delivered,780
39,C009,2026-02-08,2026-02-11,Delivered,530
40,C010,2026-02-09,2026-02-12,Delivered,600
41,C001,2026-02-10,,Cancelled,1220
42,C002,2026-02-11,2026-02-14,Delivered,910
43,C003,2026-02-12,2026-02-16,Delivered,840
44,C004,2026-02-13,2026-02-15,Delivered,390
45,C005,2026-02-14,2026-02-18,Delivered,1010
46,C006,2026-02-15,2026-02-19,Delivered,1500
47,C007,2026-02-16,,Cancelled,480
48,C008,2026-02-17,2026-02-20,Delivered,800
49,C009,2026-02-18,2026-02-21,Delivered,560
50,C010,2026-02-19,2026-02-23,Delivered,620
51,C001,2026-02-20,2026-02-23,Delivered,1300
52,C002,2026-02-21,2026-02-25,Delivered,940
53,C003,2026-02-22,,Cancelled,700
54,C004,2026-02-23,2026-02-26,Delivered,420
55,C005,2026-02-24,2026-02-28,Delivered,1020
56,C006,2026-02-25,2026-02-28,Delivered,1550
57,C007,2026-02-26,2026-03-01,Delivered,490
58,C008,2026-02-27,2026-03-02,Delivered,810
59,C009,2026-02-28,,Cancelled,570
60,C010,2026-03-01,2026-03-04,Delivered,650
61,C001,2026-03-02,2026-03-05,Delivered,1350
62,C002,2026-03-03,2026-03-06,Delivered,960
63,C003,2026-03-04,2026-03-08,Delivered,750
64,C004,2026-03-05,,Cancelled,430
65,C005,2026-03-06,2026-03-09,Delivered,1030
66,C006,2026-03-07,2026-03-10,Delivered,1600
67,C007,2026-03-08,2026-03-11,Delivered,510
68,C008,2026-03-09,2026-03-12,Delivered,820
69,C009,2026-03-10,2026-03-13,Delivered,590
70,C010,2026-03-11,,Cancelled,680
71,C001,2026-03-12,2026-03-15,Delivered,1400
72,C002,2026-03-13,2026-03-16,Delivered,980
73,C003,2026-03-14,2026-03-18,Delivered,760
74,C004,2026-03-15,2026-03-17,Delivered,440
75,C005,2026-03-16,,Cancelled,1040
76,C006,2026-03-17,2026-03-20,Delivered,1650
77,C007,2026-03-18,2026-03-21,Delivered,520
78,C008,2026-03-19,2026-03-22,Delivered,830
79,C009,2026-03-20,2026-03-23,Delivered,600
80,C010,2026-03-21,2026-03-24,Delivered,690
81,C001,2026-03-22,,Cancelled,1450
82,C002,2026-03-23,2026-03-26,Delivered,990
83,C003,2026-03-24,2026-03-27,Delivered,770
84,C004,2026-03-25,2026-03-28,Delivered,450
85,C005,2026-03-26,2026-03-29,Delivered,1050
86,C006,2026-03-27,,Cancelled,1700
87,C007,2026-03-28,2026-03-31,Delivered,530
88,C008,2026-03-29,2026-04-01,Delivered,840
89,C009,2026-03-30,2026-04-02,Delivered,610
90,C010,2026-03-31,2026-04-03,Delivered,700
91,C001,2026-04-01,2026-04-04,Delivered,1500
92,C002,2026-04-02,,Cancelled,1000
93,C003,2026-04-03,2026-04-06,Delivered,780
94,C004,2026-04-04,2026-04-07,Delivered,460
95,C005,2026-04-05,2026-04-08,Delivered,1060
96,C006,2026-04-06,2026-04-09,Delivered,1750
97,C007,2026-04-07,,Cancelled,540
98,C008,2026-04-08,2026-04-11,Delivered,850
99,C009,2026-04-09,2026-04-12,Delivered,620
100,C010,2026-04-10,2026-04-13,Delivered,720

Follow below step-by-step procedure to define custom metrics.

 

Step 1: Upload csv file to superset.

 

Data -> Upload CSV to database

 

·      Upload customer_orders.csv file

·      Specify database name and schema name where you want to upload the csv file

·      Set the table name as customer_orders

 

   

Expand Columns section.

·      Select all the columns

·      Set 'Column data types' field for this: {"order_id": "int64", "customer_id": "string", "order_date" : "datetime64[ns]", "delivery_date" : "datetime64[ns]", "order_status" : "string", "order_value" : "int64"}

 

Click on Upload button. Upon uploading the data, navigate to Datasets listing page, you can see customer_orders dataset listed there.

 

Step 2: Create custom metrics

Click on Edit action against customer_orders dataset.

 

Navigate to Metrics tab.

 

   

Click on ‘Add item’ button.

 

·      Set the Metric Key as "avg_orders_per_customer"

·      Set the Label as "AVG_ORDERS_PER_CUSTOMER"

·      Set SQL Expression as "COUNT(order_id) / COUNT(DISTINCT customer_id)"

 

Click on Save button to save this metric.

 

Click on OK button.

 

Step 3: Create a chart with this Custom metric

 

Navigate to Charts listing page and click on + Chart button to create new chart.

 

Select customer_orders dataset and Big Number chart.

 

Click on ‘Create new chart’ button. You will be taken to the chart configuration page.

 

You can see the metric AVG_ORDERS_PER_CUSTOMER metric is configured under Metrics section.

 

Drag and drop AVG_ORDERS_PER_CUSTOMER metric to the Query -> Metrics section.

 

Click on Create chart button, you can be able to see average orders per customer.

 

That’s it, you're good to go. 

  

Previous                                                    Next                                                    Home