The depth of your insights is directly tied to the quality of data being analyzed. With data coming from such a wide variety of sources in modern times, the ability to sanitize and organize your data is critical. Stratifyd allows you to enhance and modify your data through our proprietary ETL (extract, transform, and load) connector before linking it to your workspace.
Our improved ETL connector now allows for a wide range of SQL-based data transformations such as unions, and joins, adjusting column names, filtering, and performing aggregations. Doing so consolidates information into an “apples-to-apples” analytical context so you can apply one single AI-driven model to the whole group, resulting in a more complete view of the customer journey.
When should I use ETL?
Many of our customers want greater flexibility in regards to how their data looks, what it's made of, and it's core structure. This helps analysts keep the data organized, train predictive models with ease, leverage structured data fields to properly understand, and prioritize customer feedback, and more.
You may want to use ETL for simplifying omnichannel analytics. Joining data from different channels such as chats, calls, surveys, and CRM data allows you to see a true customer journey across multiple channels.
You can also utilize ETL to combine different data sources and rename fields to something that clearly identifies the contents, and deselect specific fields to remove them from your data stream.
How do I use ETL?
1. If you haven’t already, start by ingesting your data from relevant data sources by clicking the (+) New Data icon in the bottom right corner of your workspace.
2. Once you have the data streams ready for ETL, navigate to the Create New Data Stream modal once again. From the Data Transform category, choose the ETL Connector and select Next.
3. On the following screen, choose the specific data streams you would like to manipulate. Click Next.
4. Decide which operation you will need to perform on your data streams: union or join.
Union will append one dataset below the other. Typically, the data structure is very similar among the data that is being connected. For example, review data for multiple apps that needs to be combined into a single data stream.
Note: “Union All” vs. “Union” option is available in case you would like to preserve full row duplicates within your output file.
Joins are used when two datasets are intended to complement each other over a 1 or 2 common fields. For example, call center data joined with post-call survey data over a shared interaction ID.
5. Use the visual editor to complete your data strategy. You can add or remove auto-generated logic blocks that are already on the canvas. You can use the Select Columns, Add Columns, or the Filter logic blocks for even more flexibility to get you to the desired data transformation.
Once you’ve finished specifying the needed data operations, please proceed by clicking Preview Query.
Stratifyd Tip💡: The visual editor translates the specified operations into SQL, which can be previewed through the Visual/SQL toggle on the top right.
6. Review your results and click Next if you're satisfied, or go Back and make additional changes.
7. Will your data streams update on a regular basis? If not, click Next. The stream will remain static by default.
If input data streams are set to update on a regular basis, and you would like to set a schedule to have your ETL logic recur, use the toggle and then select the appropriate cadence.
8. Name your ETL stream, add any additional details such as a brief description or relevant tags and click Submit.
Please note that while SQL typically allows for storing results in a temporary table for later reference, that action is not presently available in this version of the ETL feature..
Any further questions? Don’t hesitate to reach out – we’re always here to help!