Extract optimization in Tableau - Real Business Scenario

Scenario Mode

👤 Your Role: You are a data analyst at a retail company.

📋 Request: Your manager wants you to improve the performance of Tableau dashboards by optimizing the data extracts.

📊 Data: You have a sales dataset with columns: Order ID, Order Date, Region, Product Category, Sales Amount, and Quantity. The dataset contains daily sales records for the past two years.

🎯 Deliverable: Create an optimized Tableau extract that reduces file size and improves dashboard load time without losing important data.

Sample Data

Order ID	Order Date	Region	Product Category	Sales Amount	Quantity
1001	2023-01-05	East	Furniture	450	2
1002	2023-01-06	West	Technology	1200	1
1003	2023-01-07	South	Office Supplies	300	5
1004	2023-01-08	East	Technology	700	3
1005	2023-01-09	North	Furniture	150	1
1006	2023-01-10	West	Office Supplies	400	4
1007	2023-01-11	South	Furniture	600	2
1008	2023-01-12	East	Office Supplies	250	3
1009	2023-01-13	North	Technology	900	1
1010	2023-01-14	West	Furniture	350	2

1

Step 1: Open Tableau Desktop and connect to the sales data source.

Connect to the data file or database containing the sales data.

Expected Result

Data source is connected and visible in Tableau.

2

Step 2: Create a Tableau extract from the connected data source.

In the Data menu, select 'Extract Data' and choose to create an extract.

Expected Result

A Tableau extract (.hyper) file is created from the full dataset.

3

Step 3: Apply a date filter to include only the last 12 months of data in the extract.

In the Extract Data dialog, add a filter on 'Order Date' with condition "Order Date >= DATEADD('year', -1, TODAY())".

Expected Result

Extract contains only sales data from the last 12 months, reducing extract size.

4

Step 4: Remove unnecessary columns from the extract to reduce size.

In the Extract Data dialog, uncheck columns not needed for analysis, such as 'Order ID' if not used.

Expected Result

Extract includes only relevant columns: Order Date, Region, Product Category, Sales Amount, Quantity.

5

Step 5: Use aggregation in the extract to summarize data at the monthly level.

In the Extract Data dialog, check 'Aggregate data for visible dimensions' and ensure 'Order Date' is grouped by month.

Expected Result

Extract file size is smaller and queries run faster due to pre-aggregation.

6

Step 6: Create a dashboard using the optimized extract to show monthly sales trends by region and product category.

Build a line chart with Columns = MONTH(Order Date), Rows = SUM(Sales Amount), Color = Region, Filter = Product Category.

Expected Result

Dashboard loads quickly and shows clear monthly sales trends by region and product category.

7

Step 7: Test dashboard performance and verify data accuracy against the original data source.

Compare total sales in the dashboard with totals from the original data for the filtered period.

Expected Result

Dashboard totals match original data totals for the last 12 months, confirming extract accuracy.

Final Result

Monthly Sales Trends Dashboard

+------------------------------------------------+
| Month | East Sales | West Sales | South Sales | North Sales |
|------------------------------------------------|
| Jan   | 1400       | 1950       | 900         | 1050       |
| Feb   | 1300       | 1800       | 850         | 1100       |
| Mar   | 1500       | 2000       | 950         | 1200       |
+------------------------------------------------+

(Line chart showing sales trends by region over months)

✓Filtering data to the last 12 months reduced extract size by 40%.

✓Removing unused columns decreased extract size and improved load time.

✓Aggregating data monthly improved dashboard responsiveness.

✓Dashboard accurately reflects sales trends by region and product category.

Bonus Challenge

Create a Tableau extract that uses incremental refresh to update only new data daily.

Show Hint

Use the 'Incremental Refresh' option in the extract settings and set the 'Order Date' as the key field.