It's pretty easy to underestimate the complexity of setting up a PIM process. Why?
Usually, there are vague expectations for a PIM system. A PIM is a complex software, where precise information from multiple stakeholders within an organization should be assembled in the form usable for data consumers. Different companies have different expectations and business requirements for the functionality of a PIM. That all depends on the industry, business vertical & history of a company.
Is there a roadmap? Sure. Here's something my team is using internally.
ROADMAP TO PIM IMPLEMENTATION
1. Identify the data sources and data owners inside of your organization.
It's likely your data is dispersed within your organization departments (otherwise you wouldn't be busy with building PIM, heh?:). Hence, you'll have the job of putting multiple different data streams together, aligning discrepancies in the formats.
2. Define data consumers & data end-points.
We recommend holding a meeting with your teams and everyone who is involved in your product content syndication and management processes, so you know who are the recipients of your data.
3. Learn your data and how they are provided: formats, structures, and taxonomies.
- How does this align with your data consumers' expectations?
- Which data entities are needed?
- Think of stuff like marketing texts, images & rich media, product specifications, approach to product classification & categorization, and cross-product relations.
- How your data should be localized in different languages?
- Can your data owners facilitate the requirements of data consumers? Which data transformation is needed along the way of data flow? Is data consistent enough to facilitate the requirements of data consumers?
- It's a good idea to cluster your data consumers into groups, or channels - that will help you to identify the common needs and requirements.
4. Define the roles within data flow and assign them to the right people.
Think of the data entry process, quality assurance, data enrichment, taxonomy development and management, communication flow with data consumers and data owners.
5. Define the way data hits the PIM.
Points to consider are automated data import, manual data entry, data formats, and the transportation layer (API, bulk imports, etc).
6. Define the output for PIM.
Your output means how your PIM is going to provide your data consumers with the content, in which form, and by which transports.
- Different data consumers may require different transports and formats, so consider the data format and taxonomy transformation needed for your content consumers.
- Each output variation will add up an additional complexity layer to your designs and test cases.
- It may be the case that not all data should be accessible to all consumers. All data might not be relevant to all of the consumers. Please, consider that there might be a DRM (Digital Rights Management) layer needed between the data and data consumers.
Ask Gepard Experts on How to Make a Smooth PIM Implementation
After reading this article, you shall have a clearer view of your future PIM system, data flows, and complexities behind. As well as an understanding of which parties are involved, internally, or externally in your organization. Time after time this drill needs to be repeated when there is a new data entity, data source or data consumer is appearing on the roadmap.
If you’re interested in knowing more about why companies like Amazon, GS1, HP, Dell, Rakuten, Elkjop, Lazada chose to work with Gepard, let’s get started! Feel free to ask for more details about PIM implementation challenges.