Tableau Extract Connections allow users to create a separate snapshot of data from an original data source. This extracted data can improve performance, reduce repeated queries to the original source, and support offline analysis in Tableau Desktop. Data Refresh is the process of updating an extract with new or changed information from the original data source.
Course: JustAcademy Tableau Training
Course Demo: Register for a Tableau Course Demo
A Tableau Extract is a saved snapshot of data that is stored separately from the original data source. Tableau uses its data engine to store and query the extracted information.
For example, a company may store thousands of sales records in MySQL. Instead of querying MySQL every time a user interacts with a dashboard, Tableau can create an extract containing the required sales data and use that extract for analysis.
| Order ID | Product | Sales | Region |
|---|---|---|---|
| 101 | Laptop | 55000 | Mumbai |
| 102 | Mobile | 25000 | Pune |
| 103 | Monitor | 12000 | Nashik |
| 104 | Keyboard | 2000 | Mumbai |
When this data is extracted, Tableau stores a snapshot of the selected data. If the original database changes later, the extract does not automatically include those changes unless it is refreshed.
An Extract Connection is a connection that uses extracted data instead of querying the original source for every analytical interaction.
Data Refresh is the process of updating an existing Tableau Extract with information from its original data source.
When records are added, modified, or removed in the source system, the extract may become outdated. Refreshing it updates the extracted data according to the configured refresh method and the capabilities of the data source.
Suppose an extract contains 1,000 sales records. The original database receives 100 new orders. After a successful refresh that includes those records, the extract can contain 1,100 records, assuming no other changes, filters, or exclusions affect the result.
Important: Refreshing an extract updates its data, while refreshing data-source metadata helps Tableau detect changes to fields or the structure of the source. These are different operations.
| Advantages | Limitations |
|---|---|
| Can improve dashboard performance. | Data can become outdated between refreshes. |
| Can reduce repeated queries to the original source. | Requires storage for the extract. |
| Supports offline analysis in Tableau Desktop. | Offline analysis uses the saved snapshot, not newly updated source records. |
| Can reduce data volume using supported filters and settings. | Filters or sampling may exclude records needed for later analysis. |
| Supports full and, for many sources, incremental refreshes. | Refresh options depend on the source, connector, and Tableau environment. |
Tableau provides two common ways to work with data: Live Connections and Extract Connections.
| Feature | Live Connection | Extract Connection |
|---|---|---|
| Data storage | Data remains in the original source. | A snapshot is stored separately as an extract. |
| Query processing | Queries the original source as needed. | Uses the saved extract for analysis. |
| Data freshness | Can reflect current source results when queries run, subject to caching and source behavior. | Reflects the most recent successful refresh. |
| Refresh requirement | No separate extract refresh is required. | Requires refreshes to incorporate source changes. |
| Performance | Depends on source performance, network, and query complexity. | Can be faster for suitable analytical workloads. |
| Offline analysis | Generally requires access to the original source. | Can be used offline in Tableau Desktop with the saved extract. |
| Best suited for | Cases where querying the source directly is important. | Cases where performance, snapshots, or offline access are priorities. |
Choose the connection type according to data freshness, query performance, source availability, and business requirements. Test the workbook with realistic data before deciding.
Launch Tableau Desktop and connect to a supported data source, such as Excel, MySQL, Microsoft SQL Server, or a supported cloud database.
Choose the required database, table, file, or supported data object. Review the data source page and confirm that the required fields are available.
On the Data Source page, select Extract rather than Live, where the option is available.
Depending on the source and Tableau version, the Extract Data dialog may provide options such as:
Save the extract when prompted. Tableau creates the extract using the selected configuration.
Open a worksheet and use the extracted data to build charts, tables, calculations, and dashboards.
Official guide: Tableau Help – Extract Your Data.
Tableau commonly supports two main extract refresh methods: Full Refresh and Incremental Refresh. Availability and configuration options vary by data source and deployment.
A Full Refresh replaces the existing extract data with the latest data retrieved from the original source, subject to the extract configuration.
Example: An extract contains 5,000 order records. A full refresh retrieves the current matching source records, including new records and reflecting source changes or deletions that are represented in the query result.
Advantages:
Limitations:
An Incremental Refresh updates an extract by retrieving new records based on a configured field, rather than rebuilding all extracted data every time.
For example, a sales database may receive new transactions every day. An incremental refresh can add records identified by a suitable increasing ID or date field.
Advantages:
Limitations:
| Feature | Full Refresh | Incremental Refresh |
|---|---|---|
| Data retrieved | Rebuilds using all matching source data. | Typically retrieves new rows identified by the configured field, with additional options available in supported versions. |
| Refresh duration | May be longer for large datasets. | Can be shorter when only a small amount of data has changed. |
| Existing record changes | Can capture changes included in the current source result. | Requires suitable settings or a refresh strategy that revisits changed records. |
| Deleted records | Can reflect deletions represented in the source result. | May not detect all deletions with a simple append-only strategy. |
| Typical use | Complete rebuilds and datasets with significant changes. | Sources where new records are added regularly. |
Incremental refresh uses a selected field to identify records that should be added to or revisited in the extract.
Choose a field that accurately reflects when records become available. A planned delivery date or another business date that can be far in the future may not be appropriate for identifying newly entered records.
| Order ID | Order Date | Sales |
|---|---|---|
| 1001 | 2026-10-01 | 15000 |
| 1002 | 2026-10-02 | 20000 |
| 1003 | 2026-10-03 | 18000 |
If a new record with Order ID 1004 is inserted and the selected field correctly identifies it as new, an incremental refresh may add that record without rebuilding the entire extract.
Important: A simple incremental refresh is not a universal change-tracking solution. If older records can be edited or deleted, plan a full refresh or use an appropriate supported subrange or replacement strategy.
You can refresh an extract manually in Tableau Desktop when the original source is available and the connection supports the operation.
The menu labels and options may vary by Tableau version and data source. If the workbook is connected directly to an extract file without a connection to the original source, Tableau may not be able to refresh it from the original data.
For published workbooks and data sources, Tableau Server or Tableau Cloud can support scheduled extract refreshes when the source and deployment are eligible.
Refresh schedules and available options depend on your permissions, source type, authentication method, licensing, and Tableau environment. Some sources require additional connectivity configuration, such as Tableau Bridge for eligible private-network scenarios.
Refresh frequency determines how often an extract is updated. The right schedule depends on how frequently the source changes and how current the dashboard needs to be.
| Frequency | Example Use Case |
|---|---|
| Multiple times per day | Operational reporting where frequent updates are useful and supported. |
| Daily | Daily sales or business performance reporting. |
| Weekly | Reports where weekly updates are sufficient. |
| Monthly | Periodic reports based on monthly source data. |
| Manual | One-time analysis or data that is updated only when needed. |
These are example schedules, not a guarantee that every Tableau environment supports every frequency. Configure schedules based on available options and the business requirement.
Suppose a company has 100,000 sales records and receives 2,000 new records each day.
Tableau retrieves the current matching source data and replaces the extract contents. This can capture changes throughout the dataset but may require more processing and data transfer.
Tableau retrieves records identified by the configured incremental field. This can be more efficient when the source primarily adds new records and the field accurately identifies them.
Incremental refresh may be appropriate when new records are added regularly and existing records rarely change. Full refresh may be preferable when many historical records are modified or deleted, or when a complete rebuild is needed.
For sources that allow changes to recent historical records, supported incremental subrange refresh settings can help revisit a configured period. Review Tableau's documentation for version-specific behavior and limitations.
| Problem | Possible Cause | Suggested Solution |
|---|---|---|
| Extract is outdated | No recent refresh has completed successfully. | Run a supported manual refresh or configure an appropriate schedule. |
| Refresh fails | Invalid credentials, source outage, or network issue. | Verify credentials, source availability, and network connectivity. |
| Incremental refresh misses changes | Existing records changed but the strategy only added newer records. | Use a suitable subrange or replacement strategy, or perform a full refresh. |
| Duplicate rows appear | Incremental field design, timestamp precision, or refresh settings are unsuitable. | Review the key field, source precision, and supported incremental settings. |
| Refresh is slow | Large datasets or expensive source queries. | Optimize the source query, review extract filters, and choose an appropriate refresh method. |
| New columns are missing | The source structure changed after the extract was created. | Refresh data-source metadata and perform a full extract refresh when required. |
| Scheduled refresh cannot connect | Published credentials or network access are not configured correctly. | Check connection credentials, permissions, and required connectivity infrastructure. |
| Refresh option is unavailable | The data source is Live or is connected directly to an extract without the original source connection. | Check the connection type and whether the original source is available to Tableau. |
Objective: Create a Tableau Extract and compare full and incremental refresh behavior.
| Order ID | Product | Sales | Order Date |
|---|---|---|---|
| 1001 | Laptop | 55000 | 2026-10-01 |
| 1002 | Mobile | 25000 | 2026-10-02 |
| 1003 | Monitor | 12000 | 2026-10-03 |
| 1004 | Keyboard | 2000 | 2026-10-04 |
Use this sample data in an authorized database or file. The table is illustrative and does not automatically create a refreshable connection.
After completing the exercise, you should understand how to create an extract, refresh it manually, distinguish full and incremental refreshes, and verify whether the extract contains the expected data.
An Extract is a saved snapshot of data stored separately from the original data source and used for analysis.
Data Refresh updates an extract with information retrieved from its original source according to the configured refresh method.
A Full Refresh replaces the existing extract contents with the current matching data retrieved from the original source.
An Incremental Refresh typically adds new records identified by a configured field and can use additional supported settings to revisit a range of data.
A Full Refresh rebuilds the extract from the current source result, while an Incremental Refresh is designed to update it without rebuilding all data each time.
It becomes outdated when the source changes but the extract has not been refreshed successfully.
Yes, scheduled refreshes are supported for eligible published extracts in Tableau Server or Tableau Cloud, subject to source and deployment requirements.
A simple incremental strategy may only retrieve records considered new by the configured field. Older record changes may require a suitable subrange, replacement strategy, or full refresh.
Tableau may need its data-source metadata updated. Some source-structure changes also require a full extract refresh before incremental refresh can continue.
Use an Extract when its performance, saved snapshot, offline-analysis, or refresh characteristics meet the requirements of the workbook.
Extract Connections and Data Refresh are important Tableau concepts for managing data performance and freshness. Extracts store a separate snapshot of selected data, while refresh operations update that snapshot from the original source.
By understanding Full Refresh, Incremental Refresh, refresh schedules, and common troubleshooting techniques, you can maintain accurate and responsive Tableau dashboards.
Continue learning with JustAcademy Tableau Training, or register for a Tableau Course Demo.
