PIM system implementation
To implement a PIM system successfully, attention needs to be paid to the business and the organization around the product data management process. Clear choices about processes and data integration are needed for the design of the systemWith a lack of a clear vision, organizations often struggle with implementing product information management. To avoid these struggles, it is important to ask these three questions during the planning phase:
- How to find new revenue opportunities, meet compliance measures, or increase productivity?
- What factors can improve data accessibility for better business intelligence (BI) and analytics while protecting the privacy?
- How to address cost and scalability aspects for future needs?
Note: Support from the management is crucial because PIM has impact on virtually all parts of the organization!
For a PIM implementation a methodical approach is recommended. Here you can use the accompanying seven-stage method:
The seven stages of a PIM implementation:
- Preparation: drafting plan and organization project
- Process and organization: establishment PIM organization
- System: system design
- Data: migrating existing data to PIM system and loading data
- Preparation before going live: including functional and user testing, work instructions, training and education
- Going live
- Maintain: operational people need to work with the system: ensure continuity and further optimization.
Organizing your data quality in 6 steps
Step 1- Make clear what data is there?
In order to make clear of what data is there you can ask yourself these 6 questions:
·Which data is missing or cannot be used?
·Which data is stored, but not yet in the standard format?
·Which data conflicts with each other and is not consistent?
·Which data is incorrect or outdated?
·Which data is redundant or double stored?
·Which data has not been checked for 'legal compliance'?
Step 2 – Set goals and targets for data
This step is about determining the data requirement about the “customer journey.” Deadlines and schedules are essential during this process. It has to be planned as per data per category. In order to have a more realistic and accurate planning, it is advised to perform a test for a day and will determine how much data the team can manage in one day. Based on these results schedule can be drawn up using 'extrapola- tion. You can revise this planning after step 5.
Step 3 – Design a data quality model: A model your data has to meet
In order to design a data quality model, you must first think which requirements must data meet? It’s essential to know which numbers the data should be presented in. One YouTube video format and four product images are necessary. You should also determine which languages should the data be presented in and which measurable quality issues should be used such as length and data type per data type PNG for image - xx pixel large. Lastly, it is important to know which legislation the data field should be tested in.
Step 4 – Integrate Data quality rules to monitor the data quality and make it transparent through PIM-MDM tooling
Designing clear processes means that the integration of other data overload does not lead to data pollution. Data quality rules are created that actuate workflow. Below are some examples:
AUTOMATIC DATA RULES (DQ RULES)
DQ rules ensure that relationships between products are automatically established on the basis of data and/or its characteristics.
DATA QUALITY RULES
These rules check the values in the data fields based on quality characteristics. If it has been determined that there must be three images of format A and size X, then the application detects this and can use it to trigger dashboards or workflows. Fields checking can also be done on the basis of logic from other fields, queries or artificial intelligence.
RULES THAT ACTIVATE WORKFLOWS
Automating workflow processes based on data quality ensures that a certain set of data is checked by the legal staff on the basis of regularity. After this, another set of data is checked by product specialists on the basis of technical specifications, sizes, weights, and materials used. This process, and the data that each department gets for control, can be set up via notifications, dashboards, and task lists.
DATA INTEGRATION RULES
It is, of course, very interesting to (automatically) enrich data via external data sources, various employees, and third parties. As a result, this data is brought to a higher level in a cost-efficient manner.
This also applies to the linking of data to third parties such as GS1 and ETIM. In this process, make sure that the correct data is retained! The integration of external data sources from suppliers, rich data pools or GS1 and ETIM-like parties also raises other questions:
How do I ensure that I keep the correct data?
Which existing data and the date from third parties do have to overwrite?
To organize this process, it is recommended to use automatic rules on import files!
Step 5 -Discover Exceptions and create rules and processes
Quality rules in the MDM-PIM tooling can indicate that products must always meet at least two related products and one upsell product, as well as three product images. Products that do not meet these standards must be supplemented until they meet quality standard. In order to continuously improve data quality process, it has to be checked whether the SMART criteria can be applied.
Step 6 – Monitor data quality in relation to targets
Good monitoring provides insight into the status of the products and the quality level of products. This relationship has to be regarded to planning, so management is able to use the right resources to achieve the desired data value.