Attribution

Attribution is where you manage the attributes configured for your product categories, such as upholstery material, seating capacity, or shape. These attributes are stored on your Global Products, not on individual retailer listings. So every retailer offer matched to a product is compared with the same attributed data. Attributes drive Equivalents match rules, appear as matched-attribute evidence on the comparable candidates listed on a product’s detail page, and power product filters on the Products page.

Attribution has three sections: Overview, Attribute sets, and Uploads.

Overview

The Overview tab summarizes how well your catalog is attributed:

• Attributed products. The number of products with a value for at least one attribute configured for their category, out of the products assigned to your active categories.
• Attribute coverage. The percentage of all expected attribute values that are filled in, and the total number of configured attributes.
• Missing values. The number of expected attribute values that are not yet filled in.
• Validation issues. The number of issues found in the most recent Data Quality check of your attribute data.
• Last upload. When your attribute configuration was last uploaded, and its status.

Below the summary, a category tree lists your active taxonomy categories. By default, the tree shows only categories with configured attributes somewhere in their subtree. Untick Hide empty categories to see the rest. Each category row shows:

• Products. The number of products under that category, including its subcategories.
• Configured attributes. The number of distinct attributes configured on that category and its subcategories. An attribute configured on several subcategories is counted only once, which is also why this number can be smaller than the total shown on the Attribute coverage card.
• Attribute coverage. The percentage of expected attribute values that are filled in for that category and its subcategories.
• Customer defined. “Yes” only when that exact category, not a subcategory, has an attribute you configured. A parent category does not inherit this from its children.
• Last updated. When an attribute in that category, or in one of its subcategories, was last changed.

Use the area filter to jump to a specific branch of the taxonomy, or search categories to filter by category path. A search for a parent category’s name, for example “Living Room Furniture”, also matches everything beneath it. Matching rows appear together with their parent categories. Click any category row to open its detail in Attribute sets.

Attribute sets

This tab lists every attribute you have configured for the selected category. Aperture’s built-in attribute definitions for that category are not listed here, even when they are already in effect for matching and product filters. Select a category, then review each attribute’s details:

• Attribute. Its display name and underlying key.
• Type. Categorical, boolean, text, or numeric. Numeric attributes also show their format and display unit, such as Mass with kg.
• Coverage. The percentage of products in the category that currently have a value for that attribute.
• Used by. Where the attribute is applied. “Match rules” means it can be used as matching criteria, see Match rules. “Product filters” means it appears as a filter on the Products page.

Click an attribute row to open its editor. You can change its display name, attribute group (used to organize attributes in your exports), type, and explanation, a short description of what the attribute means. It is included in your configuration exports and is used to help Aperture interpret product evidence when it automatically fills in missing values. Categorical attributes also let you edit the list of allowed values and choose one of them as a default value applied to products with no other evidence; boolean attributes can default to yes or no. Numeric attributes let you choose a format, such as mass or length; for every format except Number, which has no units, you also choose a display unit and a unit to assume when a value has none.

Before an edit to an existing attribute is saved, click Preview change to see how many of the category’s current values would fail under the new definition and how many products it would update. If the change would make existing manual corrections or other saved values incompatible, you’re asked to confirm before saving.

To add a new attribute to the selected category, click Add attribute and fill in the same fields; Aperture suggests a machine-readable key from the display name. A new attribute is available for both match rules and product filters by default.

Caution

A numeric attribute using the dimensions format (combined width, height, and depth) is not available as a product filter. Use separate width, height, and depth attributes if you need to filter on them.

To remove an attribute, open it and click Delete attribute. This removes only your customization for that category; if Aperture already defines that attribute for the category, the built-in definition keeps applying. Deleting also rejects any key, value, and pattern mappings that pointed at the removed attribute, which stay visible under “Show rejected,” and rebuilds the category’s products.

Uploads

You maintain attribute configuration through CSV uploads. Use Download template to get a starting file. If you have not configured any attributes yet, the file is a sample template with example rows. If you already have attributes configured, the file exports your current configuration so you can review or extend it.

The template’s core columns are:

• category. The full taxonomy path the row applies to.
• attribute. A short code name for the attribute.
• display_name. The human-readable label shown across the product.
• allowed_value. One acceptable value for that attribute. Add multiple rows for the same category and attribute to build its full list of allowed values.
• source_attribute and source_value. Optional columns that map a raw value observed in your source data to the matching allowed value.

To apply a file, click Upload attribution data and step through:

1. Upload. Choose a .csv file of up to 10 MiB. Validation runs automatically.
2. Review. See the file’s row and validity counts, and a full row-by-row breakdown. The breakdown shows the category, attribute, allowed value, and whether each row passed validation, with an explanation for any row that did not.
3. Apply. Available only once every row in the file is valid. Applying adds or updates the attributes named in the file and leaves attributes you do not mention untouched. Applied changes refresh affected products immediately.

Note

For an attribute that you include in the file, its allowed values are replaced by exactly the values listed in that file. Include every value you want to keep, not only the ones you add or change. Attributes left out of the file entirely are not affected.

Below the upload action, the upload history table lists every uploaded file that was validated, newest first. Files rejected outright, such as files that are not CSV, are over 10 MiB, or are empty or unreadable, are not recorded. Each row shows the upload time, row count, whether the file passed validation, and its status, which is Validated, Applied, or Failed. The Applied column shows the number of rows applied, which stays 0 until you apply the file. The Skipped column shows the number of rows that failed validation; because a file can only be applied once every row is valid, an applied file always shows 0 skipped. Where the original file is still available, its name is a link you can use to download it again. Use the pagination controls below the table to browse older uploads.

If a file passed validation but you didn’t apply it right away, its row shows a Review and apply button. Opening it reopens the same review screen with the file’s stored rows so you can apply it without uploading it again; applying at that point does not re-check the rows against your current taxonomy.