Changing a metric’s data source

In PowerMetrics, each of your metrics has a data source. Depending on your setup, this data source may be a dbt Semantic Layer, a Cube layer, a data warehouse, a data feed, or a calculation (which may combine metrics with different data source types).

PowerMetrics is a flexible solution that meets your needs today and grows with you as your data stack evolves. For example, you may start out in PowerMetrics with data that’s stored in spreadsheets or sourced from cloud services (data feeds) and later, as your data grows or your security needs change, decide to move your data to a warehouse. Or, your data may already be centralized in a warehouse and you decide to add a semantic layer for improved data modeling and management.

No matter the scenario, PowerMetrics makes transitioning metric data within your stack possible by enabling you to reconfigure metrics to use different data sources. All this without affecting existing metric visualizations, dashboards, goals, or sharing rights.

Things to note when changing data sources

• Historical data implications when switching the data source for a data feed metric. Data feed metrics store historical data that is removed when you change the data source. If you switch to use a different data feed as the metric’s data source, data collection will begin again. Stored historical data is not relevant to other metric types.
   
• Impact on metric filters, segmentations, and dimensions. If a dimension was used in a visualization but is no longer available after changing the metric’s data source, that dimension will be dropped from the visualization. If applicable, a pop-up list of “missing” dimensions can be accessed when reconfiguring the metric. You can use this information to rename the new dimensions to match the previously used dimension naming. This will ensure your original filtering and segmentation for the dimension is preserved.
   
• Instant metric limitations. Currently, instant metrics (which rely on managed data feeds) cannot be changed to use a different data source. Data sources for custom metrics (which rely on data feeds you create and manage) can be changed.
   
• Access rights. Only those with editing access to a metric can change its data source.

Lots of options

You can change a metric’s data source to refer to data from the same type of source, for example, from one data feed to another or from one data warehouse service to another. Or, you can change between data sources of different types, for example, from a data feed to a data warehouse or from a data warehouse to a semantic layer.

There are many options but they all use a similar method. As a result, by referring to the examples in this article, you should be able to successfully change a metric’s data source, no matter your situation and goals.

This article includes the following examples:

• Changing a data feed custom metric to refer to a different data feed
• Changing a data feed custom metric to refer to a table/view in a data warehouse

Example: Changing a data feed custom metric to refer to a different data feed

There are a few reasons why you might want to change a data feed metric to use a different data feed. For example, you’ve moved data from a spreadsheet to a cloud service, like moving Google Sheets data to Google Analytics or Excel data to QuickBooks. Or, you’ve moved data from one spreadsheet to another spreadsheet. Or, for whatever reason, the data feed for a metric has stopped working and you want to replace it. In every case, you don’t need to delete the metric and start over. Instead, you can switch the metric to point to the same data in a new or different data feed.

Note: As mentioned above, when you switch from one data feed to another, the accumulated history from the original data feed is lost. Data collection starts over again when you connect to the new data feed. To preserve this data make sure the history is included in the data source that will be used for the new data feed. For example, if your previous data feed included 6 months of history, make sure your data source (i.e., your spreadsheet) includes all 6 months of data before creating your new data feed.

To change a data feed custom metric to refer to data in a different data feed:

1. Click Metrics in the left navigation sidebar to open the metric catalog.
2. Select the custom data feed metric whose data feed you want to change. The metric overview opens in view mode.
3. Click the Edit button at the top of the page.
4. Ensure the Data tab is selected, click the 3-dot menu beside Data source, and select Change data source > Select a data feed. (See below.)
   
5. On the Select a data feed page, either:
   • Use an existing data feed by choosing it in the list and clicking Select. You’re ready to configure the metric.
   • Create a new data feed by clicking +New data feed and, then, clicking Select data. On the Where is your data? page, select the service to use for the new data feed. The steps to create a data feed differ depending on the service. After saving the new data feed, you’re ready to configure the metric.
6. Configure the metric.
   When possible, PowerMetrics refers to the previously-used configuration settings to auto-fill the metric’s properties (measure, dimensions, etc.). Our ability to do so depends on the data source. Sometimes, you’ll need to reselect some or all of the metric’s configuration settings. If you need help, you can find information on configuring data feed metrics here.
   Note: Under Dimensions, if you see the following note, it means one or more dimensions from the pre-existing data feed are not available in the new data source. This is often due to the dimensions being named differently. To see a list of the previously used dimensions, click the “See missing dimensions link.” Note: If you click X to close this window or you save the metric, the list of missing dimensions will no longer be available.
   
   
   Note: If a dimension's name has changed, in the list under "Dimensions", remap it to point to the previously used column by selecting the new name and typing over it with the column's original name.
7. When the metric settings have been selected, click Save metric.
   That’s all there is to it! Your data feed custom metric has been converted to use a different data feed.

Example: Changing a data feed custom metric to refer to a table/view in a Snowflake warehouse

Small to midsize companies often start out using data feeds to connect to data that’s stored locally, in the cloud, or in a data warehouse. Over time, as their data grows and their needs change (for example, they need to comply with heightened privacy or security requirements) they decide to centralize all of their data in a warehouse. To save time, instead of deleting their existing data feed metrics, they can use the “change data source” feature to transition them to direct to warehouse metrics.

Note: When you change the source for a data feed custom metric to refer to data in a warehouse, you’re also changing the metric type to a direct to warehouse metric.

To change a data feed custom metric to refer to a table/view in Snowflake:

1. Click Metrics in the left navigation sidebar to open the metric catalog.
2. Select a custom data feed metric whose source data is now located in your Snowflake warehouse. The metric overview opens in view mode.
3. Click the Edit button at the top of the page.
4. Ensure the Data tab is selected, click the 3-dot menu beside Data source, and select Change data source > See all services. (See below.)
   
5. On the Where is your data? page, in the Direct Warehouse Services section, click Snowflake.
   Tip: To quickly find the Snowflake connector, enter text into the “Search by service” field or select the tag.
   
6. If this is your first time connecting to Snowflake:
   • Click Connect account > Add a new account and enter account settings for your Snowflake warehouse. Then click Save. You can find specific instructions here.
7. If you’ve connected to Snowflake before:
   • The last account connection you used is preselected. If this is the account you want to use for your metric, click Connect. Or, select a different account connection and click Connect.
8. Configure the metric:
   • Select a table/view.
   • Select a measure.
   • Under Dimensions, if you see the following note, it means one or more dimensions from the pre-existing data feed are not available in the new data source. This is often due to the dimensions being named differently. To see a list of the previously used dimensions, click the “See missing dimensions link.” Note: If you click X to close this window or you save the metric, the list of missing dimensions will no longer be available.
     
     
     Note: If a dimension's name has changed, in the list under "Dimensions", remap it to point to the previously used column by selecting the new name and and typing over it with the column's original name.
   • Continue configuring the metric. Detailed instructions are available here.
   • When you're done, click Save metric.
     That’s all there is to it! Your data feed custom metric has been converted to a direct to warehouse metric that refers to data within your Snowflake warehouse.