How do you enrich data in a PIM system?


The sources from which all product information has been gathered usually only contain basic facts about the product such as a unique product identifier, its purchase and selling price, weight and size. The sales and marketing departments want to add more information to be able to sell them across different channels, countries, etc. More information about the unique qualities of the product, the durability for example and its ease of use can help in the selling process. In order to enrich products in a structured way all advanced PIM systems support categorization and classification.


Categorization allows organizations to sort products into specific groups allowing a better overview of the entire assortment. In addition, a classification system can help in the management of attributes. Attributes describe a specific element of a product such as its name, weight, height, length, width, etc. One of the basic concepts behind a PIM system is that it allows business users to add their own attributes to products without ICT support. In addition, management also becomes essential as the number of attributes can rise rapidly. This will be addressed separately as PIM Data Governance. With new technologies like machine learning / AI, it becomes possible to automate (bulk) classification processes. A good example is using advanced image recognition tooling that automatically generates product tags based on the image and with that auto classification is made easy. An example of this can be found at the Google vision website.

Create relationships

A commercially very interesting feature of PIM systems is the support of being able to create relationships between products. These relationships can be used for several purposes, including to:

  • Show similar products to the product shown;
  • Up-sell more expensive products;
  • Cross-sell related accessories to a product;
  • Explain that a product consists of multiple other products;
  • Offer spare parts to a product;
  • Replace an old product with a new product;
  • Communicate that it is mandatory to buy another product with a particular product;
  • Etc.

Most PIM systems in one way or another also allow users to add other media assets to a product other than attributes. Examples of media assets include photos, video, manuals, CAD/CAM drawings, etc.

These media assets are usually stored in a Media Asset Management (MAM) system which may be part of the PIM system or a separate module or system. Typical functions of a MAM system are the uploading and mass-importing of media assets, automatic assignment of media assets to categories and/or products, automatic re-sizing of photos and conversion of video for multiple distribution channels and adding meta data.

Improve data quality

PIM systems have several features to improve and monitor product data quality:

  • Automated data validation (e.g. preventing text from being entered in a field intended for a number);
  • Verification for completeness (e.g. making sure that all mandatory fields are filled in);
  • Version control makes sure revisions are kept and that the last version is always updated;
  • Data quality dashboards showing pre-set metrics and their values in a cockpit like manner. Besides this also trend reports can be displayed showing figures over time and sometimes even like for like comparisons are available.

It is key to get clean data in the system while on-boarding the Items ; in this first take one should strive to first time right models. Also using clear KPI's for the responsible staff can help. Even more if these are also part of the appraisal process. At full one should use a so-called Data Quality Framework or Model that covers all key aspects that are relevant for the company involved. The level of Data Quality in general is that this is "Fit for purpose". As an example ; all Critical data elements (CDE's) must be 100% , the next level can be OK when it's 75% and the rest is nice to have. Data Quality (needed level / defining CDE's ) is easily explained when you start from the end e.g. on your product website simulating some customer journeys.



Data access control is another feature that a PIM system must support. As multiple users enter product information into the PIM system, mechanisms to prevent users working on the same data at the same time can be a good option to have. When simultaneous enrichment is needed, logging should be possible to track and trace changes. Role based access control is a key function to limit access to specific user groups, e.g. preventing users from editing products for which they are not responsible. Finally, workflow allows the efficient organization of product enrichment. The enrichment of specific products can be assigned to specific users by a coordinator and when enrichment is done, a validation process can be initiated. A second benefit of workflow is that it makes it more transparent who has initiated which change and who has approved it. Workflow becomes essential particularly when product enrichment of the same product has to be done by different users (e.g. photographers and translators). Strong admin functionality is wanted if there are many PIM users; e.g. easy password reset options (bulk), user account management with calender options (begin / end dates). If integration with the enterprise authorization architecture is needed, this should also be possible (e.g. single sign-on via corporate apps).

Continue reading?

Share this article