Model Amplifications
Chord's data models have been built from the team's expertise in the world of e-commerce. Currently, all data flows through the same models, delivering consistent outputs across brands. However, just like the wide range of e-commerce brands that sign on to Chord, there is variety in custom data that brands want to feed into and surface from the existing Chord data models.
Chord now allows brands to extend their data by adding brand-specific mappings and settings via the Model Amplifications feature on the platform. Amplifications allow brands to structure some of their unique data inputs for modeling so they can represent metrics fitted to their business needs. Brands will see data added as additional fields in explores or see fields updated based on the configuration settings in the templates.
Once a specific Amplification is enabled, it will process through the data models in the next scheduled run.
Note that some data fields are required fields, and you'll see the following errors in case required fields are not completed. Please see an example below.
Note: Our Model Amplifications feature only accepts files in CSV format at this time. Please make sure the files you upload are CSV files.
You can also use the Download Template button, open the template in Google Sheets, and populate the rows you need to then export it as a CSV file.
You can then upload that CSV file and all valid rows will be automatically added to the configuration.
Brand operators and data analysts could use Model Amplifications to:
- Manage additional dimensions for items in the brand's catalog by adding specific attributes to their products, which will in turn show up in the data models and allow them to customize what they see and analyze on Chord.
- Reduce the need to create custom fields every time they want to report on the same dimension, like configuring the Shopify Order App IDs to a specific name.
- Add sources that are not able to be directly connected through existing data connectors. This could include tracking marketing and advertising spend for Podcasts and Radio Ads.
Each Amplification available to you can be accessed through the Model Amplifications page. The first time you configure your Amplifications, you need to fill in the Amplification templates with your custom data mappings.
Once you have filled out each Amplification template with your custom data mappings, you can then enable them by requesting activation (templates that do not have any data will not be enabled) .
It will take about 48 hours for your Amplification to be enabled and for your custom data to be surfaced in your existing Chord data models.
Once the 48 hour activation period has taken effect, your Amplification will be enabled and will be available in your Chord data models. The activation process happens only once, adding additional data mappings to already enabled Amplifications will continue to automatically feed into your Chord data models.
Going back to the Model Amplifications page, you will be able to see which Amplifications you have enabled, which are in progess, and which still need to be enabled.
Data Upload Limitations
- Users cannot upload files greater than 10MB. Please ensure that your file upload doesn't exceed 10MB in size. This translates to roughly more than 30K rows of data.
- Please only upload CSV files. The Chord platform will not be able to read any other file format for model amplifications.
- Please do not remove the "Key" column provided in the CSV template. This column is used for internal purposes and should not be altered.
The Chord data platform also supports bulk upload of data, which in turn supports brands that need to upload many rows of data, or need to need to add thousands of new SKUs to a specific data model at any given time.
Note: Once you bulk upload data to our amplifications feature, there may be a wait time to view the results on our platform. This is governed by the data refresh cadence set for your organization as well as the timing of the last data sync with Fivetran.
To bulk upload data, you can follow the steps below:
On a given model amplification template, download the template, fill in the desired columns with your data, and click the Upload CSV button on the template page. From here, you can attach a CSV file to upload.
Once you attach a valid CSV file, the CSV upload modal closes and you will see a message saying that the file is being processed.
During processing, the Chord data platform will check the new file and validate all rows in the file to ensure that there are no duplicates (i.e. check required columns and uniqueness). Once validated, the information in the new sheet will be appended to the original file.
All valid rows of data in the uploaded CSV file will be appended to the model amplification. Invalid rows of data will be provided in a new file that can be downloaded by the user to reconcile once the initial upload processing is complete.
- Question: I need to update a previously uploaded / provided datapoint in an existing amplification. Will the model amplifications feature find and replace this information when a new sheet is uploaded with current information?
- Answer: The model amplifications will simply append the updated data as a new row to the sheet. Existing rows will neither be removed nor updated based on what’s provided in an uploaded CSV. In other words, model amplifications will not "find and replace" one datapoint in a pre-existing row at this time.
- Question: What is the best way to update a large number of rows via an upload without changing already uploaded data?
- Answer: If there are a large number of rows that would be too time-consuming to update manually (i.e. row by row directly in the Amplification UI), these rows can be updated via the Upload CSV functionality.
- To prevent unnecessary modification, you can:
- Download the amplification data in its entirety.
- Make the necessary changes directly in the CSV.
- Clear all of the data in the amplification and then re-upload.
If you have any additional questions or require additional help, please reach out to [email protected]!
Listed in alphabetical order:
Template Name | Fields Included | Notes | Release Date |
|---|---|---|---|
Required fields: source, date, spend | This configuration enables you to include advertising spend for sources that are not able to be connected directly to data connectors, such as podcasts, newsletters, etc. | July 8, 2024 | |
At least one channel plus the condition | This configuration maps user-defined Channels and Sub-Channels to page utm and referrer details. | July 12, 2024 (select tenants) | |
Required field: Campaign Name | This configuration maps the campaign name to different attributes. It allows for only one row per campaign. | October 25, 2024 | |
Order App ID | This configuration maps the order app id from Shopify to a user-defined name. | July 8, 2024 | |
Order Filter | At least one condition is required | This configuration applies specific conditions to create a flag for orders in the Orders, Line Items, and Sales Explores. This configuration is designed to flag orders you want to filter out. | July 8, 2024 |
Required fields: Date and Forecast Version Name
| This configuration allows Chord Performance brands to upload their daily planned forecasts of gross and net revenue, along with number of orders and number of customers. | January 22, 2024 | |
Required fields: SKU, product name, variant name, and existing hierarchy fields (product type, product_sub_type) | This configuration maps additional attributes per variant that may not be available in the OMS, including hierarchy, size, function, and color Please note that this configuration only allows for one row per variant. | July 8, 2024 |