PIM-first or ERP-first? The right strategy for creating product data for successful projects

An example
A wholesaler buys in goods from various manufacturers and sells them on to various customers. Manufacturers supply product data in varying quality and formats. The decision on which products from which manufacturer make it into the portfolio, and which new manufacturers should be added to the range, is driven by terms and commercial considerations. Further enrichment with product information for the wholesaler's own customers is carried out solely for the adopted articles, in a downstream step.
At first glance, everything seems clear: products start life in the ERP system, born from the master data of imported price lists.

Let's assume, for example's sake, that each manufacturer uses its own interpretation of format for supplying data to our wholesaler, and that data quality in places still holds great optimisation potential.

PIM systems have specialised tools for importing and exporting differently structured information into their own data model. Their validation and quality assurance mechanisms can, for example, also be used to ensure that only data in a clean format with a sufficient level of maintenance is provided to the ERP.

Then importing manufacturer data via the PIM could make the setup and maintenance of many ERP interfaces – and the error handling that's especially costly in ERP systems – unnecessary from the outset.

When deciding which system takes the lead in product and article creation, there are a few guiding questions that help with good design of the later process:

• How do the first product and article data reach the company? 

  • If they are created in-house, whether through own product development or manufacturing, or because suppliers do not deliver product information usable in automated processes, several signs point to a PLM or PIM system being a good starting point for data capture and maintenance.

  • If sourced from external systems, ERP can be good starting point, provided it has well-configurable interfaces with efficient exception handling, and provided external data sources deliver stably in format and content so relevant data can be retrieved specifically. Otherwise, one should weigh whether PIM system with its pronounced ETL mechanisms should be placed upstream to establish and safeguard quality of data to be imported into ERP.
     

• Is third-party product data used to support decision-making or for tiered distribution?

  • If data on competitor or not self-distributed products available in good quality, strategic conclusions often drawn from that, whether and to what extent own range complete, where diversification could pay off and how own items stand compared to third-party items. Since such data usually should not burden ERP system and PIM systems bring mechanisms for product comparisons out-of-the-box, they interesting entry point for such scenarios.

  • In scenarios where items beyond one's own range are offered and only created in the ERP system upon order (particularly in wholesale and intermediary trade there are a number of scenarios where the sheer volume of potentially interesting items can hardly be managed by an organisation), a PIM system with the existing manufacturer data can enable orders and reservations; the shop or ordering system then creates a sales item in the ERP as part of the order, using the master data available in the PIM, so that the purchasing and sales processes only need to deal with the items actually actively managed.
     

• Is there externally driven product design, or customer-initiated article creation?

  • Particularly in B2B scenarios such as make-to-order manufacturing with a high degree of customisation, it can make sense to prepare the customisation in a PLM or PIM system and only pass the fully designed products and articles to the ERP, in order to avoid mapping all possible combinations, which grow exponentially with the options, there from the outset.

  • If systems deliver data on requests that cannot be satisfied with the existing product master data, they can be systematised and collected in a PIM system to support product planning processes and decisions on range extensions.
     

• Can existing processes be streamlined or made more efficient? Are there recurring bottlenecks in ERP master data maintenance?

  • In many companies, access to data maintenance in the ERP is sensibly heavily restricted, so the relationship between product initiators or product managers and the data-maintaining staff often leads to delays in the product creation process. A PIM system can significantly widen this bottleneck by pre-maintaining information in the PIM system and making it available to the ERP via interfaces, so that only missing ERP-specific master data needs to be added in the ERP system. This enables, on the one hand, a higher degree of parallelisation, and on the other hand relieves the ERP staff, so that overall the time-to-market is optimised.

  • As a rule, information for creating product master data is not generated by the employees who maintain it in the ERP system. It is usually pre-captured using various tools and then transferred to the ERP system manually or semi-automatically. Systematic pre-capture in the PIM, combined with validation and data quality assurance measures, can often avoid redundant process steps that already exist. In addition, transmission errors are ruled out.

  • Workflow mechanisms in the PIM can control creation processes so that, through sensible parallelisation and defined quality gates ensuring the data and information needed for the next step are present, loops, queries and waiting times can be minimised.
     

• Are there compelling technical or business-strategic reasons to locate the creation of products and articles in a particular system?

  • Some systems – particularly those used and continually adapted over a long period – dictate certain process flows and make switching to a leaner process difficult and/or uneconomical. The risk of a process change, bearing in mind the natural inertia of long-used working methods, should also be examined.
    In such scenarios, it makes sense to pursue ERP and PIM systems with an ERP-first approach combined with a side-by-side approach, in order to gradually redistribute functionalities and processes within a planned change process and slowly and gently rebuild the IT architecture.

  • Even with a planned ERP replacement by a newer system, this approach lets you use the PIM as a bridge between the old and the new system, whereby product creation follows a pattern ERP[old] à PIM à ERP[new], and the old ERP is eventually phased out.

Conclusion

How this process is ultimately shaped optimally is highly individual, as it depends on technology and business model, existing approach, company culture and organisation, the standing of individual business units, and possibly even individual products.

The guiding questions and background considerations outlined show that the change processes accompanying a PIM implementation, together with the new possibilities gained from it, offer many options for questioning, streamlining and optimising existing workflows. They are intended to help ensure that decisions on changing operational workflows connected with a PIM project are made consciously, without bias and from various angles.

Well-designed product creation and maintenance process basis of many business cases that draw on product data, but in turn also influence mapped portfolio and sales and communication measures. It should avoid loops, enable parallel work steps and adapt dynamically to company strategy and market requirements, while at same time ensuring data quality and protection of processes running in ERP system and other business-critical systems.

Want to introduce a PIM system at your company? We'll help you choose and implement the right solution for you! Feel free to get in touch.