The foundation for the PIM system: analysis and planning

Aspects and topics that you should consider before starting the project

A PIM system should fit seamlessly into the existing landscape. By landscape, we mean your existing IT systems, tools and organization. And in particular the data model for mapping the product logic. Standard software is good, but only if individual aspects can be configured with it to ensure optimum integration. Knowledge of all PIM-relevant aspects should be acquired at the beginning and not "as we go along" during the course of the project. Of course, not all details need to be known, but the basic structural relationships must be on the table for a good concept and reliable planning at the start. The documentation prepared then serves as a guideline for the project.

Checklist for analysis and planning

Typical questions that arise for analysis and planning are

1. which topics should be addressed?
2. who should participate?
3. how much time is required?
4. what specifically is the outcome?

Question 1: The list of topics for the agenda

There are cross-cutting topics that must be dealt with in every PIM project and there are individual topics for each company.

The list of topics should be agreed in advance in an agenda so that the topics can be dealt with in a targeted manner:

• ERP interface
  Is it a "lean" interface where the logistical and commercial article master data is transferred or a comprehensive interface with technical product data, packaging data, configuration data, etc.?
• PIM data model
  This relates to product structure, product relationships, classification and data sovereignty.
• Content data model
  This concerns the categorization of images, documents, videos and texts. The image workflow with image conversion, image sections and formats. Keywording, metadata, etc. In an expansion stage, a text library for text modules, text generation with templates or via AI and a media service portal.
• Translation process
  What exactly does the process look like, should AI be used for pre-translation, should it be checked and corrected online in the countries, should a translation agency and TMS be connected?
• Initial data transfer
  Is there a migration from an existing PIM system or import from other existing data sources? How well can this data be used and to what extent does the data need to be adapted?
• Channel output management
  Which outputs or rejections are relevant. These are controlled by means of catalogs, which can be configured with language/country, product range, price list, output structure and definition of content.
• Data provision via API interfaces
  For example, store, website, configurator, CAD portal and mobile apps.
• Data provision via e-catalogs
  Including BMEcat and Excel exports for trading partners or specific XML or JSON formats for portals.
• Read access
  The PIM is a valuable source of product information, especially for sales.
• Print production with InDesign
  Be it catalogs, price lists or data sheets. The relevant media must be checked in terms of layout, structure, data origin, language changes and automation.
• Graphic production with Illustrator
  If language-specific graphics and dimensioned drawings are also to be automated, the procedure must be checked.
• IT system
  Should the system be installed in the cloud or on premise? Typical questions concern the system parameters, access with SSO, the security architecture, backup and maintenance intervals.
• Users, tasks and authorizations
  User behaviour, user groups and necessary workflows can be derived from this.
• Planning
  Project organization, timeline and effort.

Question 2: The participants

The group of participants should be a good mix of specialists and decision-makers: The participants consist of your responsible employees and decision-makers from marketing, IT, graphics, product management, sales, e-commerce and engineering. Typically, there are 5 to 15 participants, who are divided into 3 to 5 teams. The project manager is always present, all other participants are invited on a topic-specific basis.

Question 3: The agenda

1-2 hours can be planned for each topic. Depending on the topics, one or two, sometimes up to three workshops are necessary. For cross-cutting topics such as the PIM data model, almost all participants are on board. In the case of print, for example, only the graphics team and in the case of the IT system, only the IT team. Also take into account the preparation and follow-up time. Some data must be compiled in preparation and the documentation must be checked afterwards.

Question 4: The result

The documentation must present what has been discussed in an understandable way. The overall picture must be comprehensible. All requirements must be highlighted and it should be clear what can be done and how it can be done. The quantity structures should be worked out so that the data can be estimated numerically.

Ideally, the tasks for the employees should also be specified so that the internal resources can be estimated. After all, a PIM project is teamwork between the in-house team and the provider's team.

Finally, there must be a concrete time and resource plan, with installation, workshop and training dates and the corresponding costs. These milestones should also be planned: When will the ERP interface be up and running? When will the system be live? When will which releases be productive? The first success is achieved when the system goes live, but the really important steps are the rollouts (store, website, e-catalogs, print media).
Ultimately, the result is used to prepare an offer and make a decision.

Conclusion

Don't start the project without a sound analysis and planning. Nobody would want to build a house without a plan. The effort for this first preliminary project phase amounts to approx. 5% of the total effort and will, I can guarantee, be very worthwhile.

And let us show you the functionalities of the individual topics using a prototype - with your data! This makes it much easier to understand the connections "click by click". And with standard software, it should be possible to build a prototype with manageable effort.

This approach allows you to get to know the provider very well: How good is he at listening and understanding? How suitable are their solutions? How clearly do they communicate? How well documented are they? How do they moderate and lead? And of course, what is the chemistry like?