Design considerations for your data

  • Updated

You should think about these considerations when designing and organizing your product data before you start your Optimizely Product Information Management (PIM) configuration. This will save you time and simplify the implementation process.

General Governance

  • Define your PIM team – You should identify your key members early in the process who will help you get your data in place. They can begin collecting any data or assets you need. 
  • Keep data clean – You should identify any inconsistencies in your data, such as misspelled words, irregular capitalization or abbreviations that someone might not understand. Clean data is very important to the integrity of your products. You can edit your product information in the PIM, but it may be easier to edit the data in your current system or spreadsheets.

Property Groups and Properties

Set up your property groups and properties to have a successful experience with the PIM. Here are a few governance rules you should keep in mind before you define your properties:

  • Give specific names to properties – Property names must be specific and unique, but the display names can be shared. For example, your properties could be Faucet_Finish and Countertop_Finish, both with a display name of Finish.
  • Create property groups by product family – If you have properties that run across multiple products, such as color and style, create separate property groups by product family and then add their respective properties specific to each. For example, create separate property groups for Faucets and one for Countertops, then add Faucet_Finish and Countertop_Finish to the respective group.
  • Create specific properties for easier attribute filtering – Separating property groups allows your users to filter based on the properties of that group. For example, you may have chrome or nickel options for faucets but matte or glossy for countertops, even though both are part of the Finish property.
  • Think about your site pages – Consider where properties that are attributes land on the Product Detail page on your site. 
  • Think about product templates – You may want to have product templates to make editing product types easier.

Product Templates

You can pull the property groups with properties you created into product templates.

  • Think about your product families – Design product templates by product family. For example, a merchandiser may create a Shoe product template that contains essential data elements for that type of product, such as color and style.
  • Adjust properties by template – You can create separate templates for Athletic Shoes and Dress Shoes. If you have a shared Shoe property group, you can include this on both templates and then adjust required and recommended properties within the property group for each template.

Category Trees and Categories

Category trees and categories are the overall taxonomies for a site and your site navigation. 

  • Think about how your customers navigate – Categories are how users navigate your website. Categories can be regular product categories or parent categories for other (child) categories. Parent categories display a category list page, while product categories display a product list page.
  • Avoid too much depth – Too many levels of categories can cause navigational issues on your site and can be cumbersome for your users. You should stay within three levels in your category structure.


Organize asset folders before you load any assets.

  • Be specific with folder names – When planning how to organize your digital assets, avoid creating a traditional file structure with vague folder names such as Images or Specifications.
  • Follow a structure – Think about your product categories, brands or documents you maintain.
  • Use subfolders – Use subfolders to further organize assets. 
  • Consider auto-assigning – Use an asset name with a product number to auto-assign assets.


  • Create import templates – Use an import template if you are getting data from specific suppliers in a consistent format.
  • Do not worry about a consistent spreadsheet format – Because you are creating your own templates, you do not have to modify the spreadsheets before importing if you column headers (source fields) map to target properties in the PIM.
  • Resolve errors after import – Data does not have to be perfect. If some data is not mapped properly and produces import errors, you can resolve them after the import. The system imports products without errors and isolates those with errors.