Intro
This feature allows you to import your business data (e.g. margins, sell-through rate, return rates, etc) as a way to improve your Product Inventory with additional information. This is useful to optimize towards business goals such as profit. You can then apply your custom Product Attributes in Product Insights to filter for individualized data insights. This is also useful for purposes such as creating your own Custom Metrics.
The process of connecting custom Product Attribute data uses either one of two sources:
A dynamically updated spreadsheet file (ideally once every 24h)
Preexisting internal feeds from Meta, Google, or custom sources
Multiple custom Product Attributes can be created from the same source.
Requirements
Before you start, confirm you have three required materials (different depending on your source):
1) An active Market with Inventory in ROIH | 1) An active Market with Inventory in ROIH |
2) A source of your custom data:
| 2) A source of your custom data:
|
3) An upload method of your custom data:
| 3) An upload method of your custom data:
|
For any questions or problems, please contact ROIH Support for help with the proper set up process.
Integration
Navigate to Settings > Markets and Assets:
Select an existing Market, and select the "Product Attributes" tab:
Click the "Connect" button on the right side to open the Import source file panel:
Select the appropriate Upload Source Type based on whether you're connecting a spreadsheet file or sourcing a preexisting feed:
FTP if your file URL starts with “ftp://
SFTP if your file URL starts with “sftp://
S3 if you use Amazon S3 bucket
Meta Catalog for any connected Meta data
Google Merchant Center for any connected Google data
Inventory Feed for any connected custom data
Fill in the associated entries:
For an uploaded spreadsheet, Name and URL (if the file is password protected, also fill in security details)
For the custom data source (such as Meta or Google), the respective Meta Catalog, Google Merchant Center, or existing Inventory Feed
Click the "Next" button
If you are unsure the file was successfully uploaded or want to confirm validation, please contact ROIH Support.
Continue with the pairing setup:
Pairing Field: Select the column from your source file the data should be paired with. Conventionally, it is probably ID or Group ID.
Product Hierarchy Level: Select the field in your inventory with which the provided data should be connected. Ideally, select one of the levels that is the same as the type of the ID in the provided file linked above. You can check the examples to find which value is in the same format, or check which fields are used as Product Hierarchy Levels.
Recommendation tags will be displayed next to the most appropriate source based on automatic calculations:
Define Product Attribute(s): Select the first product attribute you want to use, and fill in “Name” (and optionally “Description”).
Value Type: Used to identify what type of data the desired attribute contains. They are shown according to the platform.
Integer: Whole numbers with no decimal points (e.g. -3, -1, 0, 2), for use with measurements like the number of items sold, stock levels, etc.
Float: A decimal number with floating-point precision (e.g. 3.14, -0.5, 2.0) for use with calculations that require more accuracy than integers.
Money: Currency units for use with measurements like revenue, spend, etc.
String: A sequence of characters that can contain letters, numbers, symbols, and spaces for use with measurements such as identifying hero products, high or low margin buckets, any other custom labels, etc.
Percent: A portion of 100 represented by the symbol "%", for use with calculations not in absolute units such as sell-through rate, net margin, etc.
Aggregation: Defines how the data is shown on higher levels (e.g. product variant → product → category). If you upload the margin on the product level, you would probably want to average margins on the higher (category) level. For a total amount of returned items, you would like to ideally see the sum for all returned products in the category. If you have Margins in buckets (i.e. Very Low, Low, Med, High), CONCAT combines all values together, which can be used to create filters such as “Category containing Low products”. For more information, see the Example of Correct Aggregation for String Value Types below these steps.
SUM: Adds all of the values together
AVG: Calculates the average of all the values
MIN: Returns the lowest value (this can be numerically lowest or alphabetically lowest based on the Value Type)
MAX: Returns the highest value (this can be numerically highest or alphabetically highest based on the Value Type)
CONCAT: Concatenates (combines) all of the values into a single compound value
To connect multiple attributes from the same source, click "Add Product Attribute" and repeat the configuration steps
Click "Save" when finished, and the connected Product Attributes will be available to view in the panel
Example of Correct Aggregation for String Value Types (click to expand)
Example of Correct Aggregation for String Value Types (click to expand)
Consider how ranges of margin may be represented by a string variable: high, medium, low for three different T-shirts in multiple variants based on sizes.
The first T-shirt has mixed margins because some sizes are popular while others that may be extremely large or small are discounted since they are not selling that well:
1. T-shirt XS size has very low margin
1. T-shirt M size has medium margin
1. T-shirt L size has high margin
1. T-shirt XL size has very low margin
The second T-shirt is new so it has high margin on all sizes:
2. T-shirt S size has high margin
2. T-shirt M size has high margin
2. T-shirt L size has high margin
2. T-shirt XL size has high margin
The third T-shirt is almost sold out and needs to be cleared so it’s discounted on all remaining sizes
3. T-shirt S size has low margin
3. T-shirt M size has medium margin
3. T-shirt L size has very low margin
The result in Product Insights would be the T-shirt as a Product Group where the unique values of each variant are concatenated, separated by “;” and ordered alphabetically:
1. T-shirt - high; medium; low; very low
2. T-shirt - high
3. T-shirt - medium; low; very low
Now, suppose you need to create a segment with high margin products. So, when you go to margin filtering in the attribute filters, you will be able to pick from all possible combinations. In the case of our three T-shirts it is:
high; medium; low; very low
high
medium; low; very low
Depending on the definition of the “high margin” segment, you have more options of setting them up, for example:
Use the margin attribute filter contains “high” and then get all products where at least one variant is of high margin (weak high margin) which results in T-shirt 1 and 2
Use the margin attribute filter equals “high” to get only products where all variants have high margin (strong high margin) which results in T-shirt 2 only
Or: alternatively, revamp the margin value into more granular ranges such as very high, high, higher-mid where the contains “high” would cover a broader spectrum of the high margin products, depending on how much “high” you want it to be
Usage
Navigate to Product Insights
Click “Add first/next attribute” in the filtration panel to display a list of all possible Product Attributes. Custom Product Attributes are identified with a 🔗 chain link icon:
Select any custom Product Attribute and click “Show additional info” to see “value type”, “aggregation” and “data source”. This helps you to figure out how to configure the filter.
Selecting mutually inclusive criteria for a Condition is possible, and the number of products that are being filtered in is displayed at the bottom of the Attribute selection panel. This is an estimation from all products (even those out of stock or deleted).
FAQs
Q: Can I add attributes just to categories?
A: No, you can only pair attributes to product variants or products. However, their values are aggregated through categories.
Q: How often do you recommend to update the source file?
A: ROIH updates the business data once per 24h.
Q: What if I have business data like this in the main product feed already?
A: This is also possible, but please contact ROIH Support.
Q: What if my margin changes in time, can I see it in charts?
A: Product Insights offers the ability to view trends or breakdowns for metrics only. As Margin is considered an attribute, you can't analyze its trend in the Trending or Breakdown charts.
We strongly encourage you to use your margin data and create calculated metrics with it (such as POAS or Gross profit from your ads). This will allow you to see trends in the metrics related to margin. Read the article here for more information.