ERP master data: “Taming the beast”

By John Fatkin, Lead Consultant – Supply Chain Management on
Let’s face it: master data doesn't usually precipitate loud applause and excitement at conferences, nor is it cited as the reason to justify an ERP project. Sooner or later however, we learn of its influence on our business processes and the inevitable responsibility that we have to be, not just custodians, but controllers of its health and relevance.

Somewhat begrudgingly, the task of master data maintenance is accepted, but a lack of commitment and tardiness to keep the information accurate, complete and up-to-date can allow those legions of tables, fields and characters to grow chaotic. We realise with remorse that we need to quickly find a way to tame the beast that master data has become!

Creating the beast

It starts in the blueprint design. Unwittingly, we can set ourselves up to fail by chasing a desire for detail in reports. In those whiteboard sessions, we construct the architectural tenets of the ERP solution, far from the actual shopfloor and factory offices where the effectiveness of the design will become starkly apparent. Design decisions are driven by the sparkle of core business process outcomes. We marvel at the colourful planning tools, the easy drag-and-drop order functions and the interconnectedness of our actions upon the entire supply chain. The reliance on the underlying master data to enable such processes to operate however, might not rate a mention on the blueprint documentation. What can help?

  • Design master to be simple and with maintenance in mind data: In many areas of master data, SAP provides opportunities to do master data simply and avoid duplication. For example, Task Lists (Recipes, Routings, Inspection Plans and Maintenance Plans etc.) can be created to allow a single List to support multiple Materials. Surprisingly, many businesses persist with poorly justified, individual Lists. Further, Variant Configuration demonstrates the best example of minimum master data to manage a multitude of product and process variations - a single Bill of Material and a single Routing supported with logic rules can, not only avoid the maintenance of dozens of BOMs and Routings, but operate with much greater consistency. To further help with consistency and reduce maintenance, 'Profiles' allow groups of master data fields to be maintained efficiently. For example, the MRP Profile allows fields and field values to be selected and maintained as a group. Automatically, all Material Masters that reference the MRP Profile, get the benefit when a change is made.
  • Don't over-react to master data: Having heard rumblings of master data woes from others, some businesses approach an ERP design with a proud aversion for master data. These are the people who prefer to raise Purchase Orders with free-text, rather than to use Materials. It's a pity that they won't know their annual spend or use knowledge of their future demand to get a price discount from their suppliers. Remember, there is power in master data.

Feeding the beast

After go-live, the likelihood of successful master data management will be evident within the first couple of weeks. If we adopt a sluggishness towards our master data health, the beast will start to grow. It's vital to be proactive in these matters. What can help?

  • Devise slick master data maintenance processes: From day one, master data maintenance processes should be ready to operate. This means having overt agreement on the responsibility to maintain each master data object and turnaround time - even down to field level if required. As master data of ERP systems is often cross-functional, there is a high risk of delays as contributions need to be provided and interdependencies processed. The process status must be given high visibility so that blockages can be promptly managed. For this reason, master data dashboards and workflow are assumed tools. The old approach of attaching forms to emails, sends problems to ground and becomes a frustration to chase.
  • Use tools to support changes easily: Using the standard 'change' transactions is fine for first principles, but can be a tedious method for daily tasks.  In recent times, SAP has responded to the plight of data entry and the high number of keystrokes that can be involved. Screen Personas are a great tool to allow a user to design their own transactions and radically reduce the keystrokes for recurring tasks. SAP Engineering Change Management and Mass Maintenance transactions offer an efficient approach to large scale changes and the synchronisation of related objects. Where co-ordination of changes is highly complex, then SAP Management of Change is an efficient tool. In addition, third party software providers such as Winshuttle, have continued to provide focused tools to ease the effort of large scale master data changes.
  • Train people: It sounds obvious, but months after the ERP implementation, the natural transfer and attrition of staff can leave the team depleted of the skills and awareness needed to manage master data. This is a vulnerable state. Act to train back-ups and cross-train existing team members into other modules. An under-informed and under-resourced team will provoke the beast to grow!

Taming the beast

Even when the ERP system has settled down there is a need for vigilance to keep the beast at bay. So what can help?

  • Measure it regularly: The Master Data Maintenance process can have KPIs on data accuracy and timeliness. There can also be root cause of delays to master data creation and change. Then there is the quality of the data itself – ie. number of missing fields, the incorrect values and the number of inconsistencies between fields. Don't let the operation be treated like a black box. Corrective actions are enabled by visibility of the problem first.
  • Select your review target wisely: Nowadays, reviewing master data might seem to be a luxury. Coupled with the enormous amount of data that exists, there can be cause for consternation as to how a proper master data check can ever be achieved. Answer? Don’t check everything, every time. There is a priority to where time should be spent. Within most areas of master data, there are the records that haven't been transacted within the last ‘x’ months. Then, there are the records representing little dollar value. By a process of segregation, it becomes apparent which master data records are most influential on cost, value and customer or supplier relationships. These are the records that deserve most attention for review and correction.
  • Clean it: Despite the best efforts and diligence, data errors will occur. There's more than just correcting an obvious error too – screening should be in place to detect and correct duplicate records (eg. John Smith, J Smith, J.B. Smith etc.).
  • Cull the dross: As the mountain of data grows, there are obvious maintenance advantages to retiring old data records. Note, old materials numbers, plant numbers etc, should not be re-used if any historical records were made with reference to them, otherwise, a corruption from the associated records can occur.
  • Be alert to division in the camp: At all times, there should be an awareness to the tangible and intangible clues regarding master data health. As master data maintenance in an ERP system is a team task, any team members who lack visibility of the process or experience the repercussion of errors from others, can lose faith in the process and withdraw their support. In such cases, a weakness in the process can extend quietly and affect others. Hence, process owners need to intervene and resolve such issues promptly with appropriate insight and understanding.

Finally, it’s indisputable that accurate and timely master data is essential to the operation of an effective ERP system. Business management would do well to place high value on the role of its maintenance. Ultimately however, this advice won’t actually eliminate the beast of master data, but it might help to show it "who's boss"!

For further information please contact us