de
This article is part 1 of a three-part article series on the topic of "Master Data".
Without doubt an ERP system can only be as good as the master data quality it possesses. Even if this topic is already an ongoing issue, the rapidly advancing digitalisation threatens to catapult it from the frying pan into the fire. There is a danger of being caught between the millstones of the flood of data and the simultaneously increasing demands on data quality.
Master Data: An Ongoing Issue That Needs to Be Tackled
The numerous problems with master data have already been discussed several times. They involve duplicates as well as incomplete or outdated information. But the effects of bad master data are usually not immediately noticeable, the consequences are much more insidious. Over time, employees then have more and more queries, and the coordination processes become much longer. This creeping effect causes the staff's trust in the company's own data to dwindle in the long run. It goes in the same direction as the often quoted frog metaphor, where the frog - sitting in the pot of slowly heating water - no longer manages to save itself in time. If decisions have to be made under deadline pressure - not uncommon in everyday work - employees learn to help themselves. Under certain circumstances, incorrect master data are simply accepted or gaps are filled by more or less arbitrary estimates.
The New Challenges of Digitalisation
As if the general master data situation wasn't challenging enough, the ongoing digitalisation poses even more challenges for companies. The amount of data continues to grow rapidly. At the same time, the quality requirements for the data are also increasing.
Data Flood Due to Diversity of Variants
The trend goes right across all industries: From the food industry and furniture manufacturing to mechanical engineering and automotive - customers increasingly want more individual products and more often in smaller delivery quantities or sizes. In the service sector, too, customer wishes are becoming more and more differentiated. Technological progress favours this trend by making more and more variants technically feasible and economically affordable. A suit or jeans made to measure: today, production can be initiated by entering the individual sizes into an online shop form. Individual parts in small series can be ordered by uploading a CAD drawing to the corresponding supplier websites. For master data in ERP systems, this means one thing above all: their number is increasing massively due to new customers and products.
3D Printing Favours Data Growth
Technologies such as 3D printing are also driving this. This allows physical objects to be produced by applying layers. This makes it possible to produce individual parts and also small series - there is often talk of batch size 1 - economically. This also leads to a strong increase in the number of product variants, and with them the (master) data volumes continue to grow.
Data Flood Through the Internet of Things (IoT)
In addition to the diversity of variants, the Internet of Things (IoT) is causing a data explosion. More and more measurement parameters can be recorded more sensitively and at the same time more cheaply by sensors. In addition, the costs for data storage are constantly falling. This in turn means that more and more machines, devices and other appliances are being equipped with sensors and generating data. These include, for example, aircraft engines, lift and train doors or food in transport and storage containers, whose cold chain is monitored by sensors around the clock. In addition, the development of providing many devices, machines and other components with a digital twin - i.e. a digital copy of the physical object - is leading to a further massive increase in data volumes. Master data is also increasing and forming more and more links to other data.
The new technological possibilities and the increase in data then cause the need for further solutions - which in turn also generate new data - to rise. These include solutions in the areas of data protection and data security. At the same time, it is becoming more and more difficult to keep track of master data and other data, and the maintenance effort and the susceptibility to errors continue to increase. In the product data area, for example, new products, variants and versions as well as new hierarchy levels for master data are being added.
Digitalisation & Increasing Demands on Master Data Quality
At the Same Time, Digitisation Raises the Standards
However, another trend taking place at the same time is that digitalisation is increasing the demands on data quality. For example, automation and machine learning will play a major role in managing the flood of data. However, these technologies require high quality (master) data that is as accurate and complete as possible.
Product Life Cycles Are Becoming Shorter and Shorter - Keeping Data Up to Date is Becoming More Complex.
Product life cycles are also becoming shorter and shorter due to digitalisation and new technological possibilities; in addition, niche requirements are also increasing. For master data, this in turn means that their lifespan is becoming much shorter and they become obsolete much more quickly. They therefore need to be updated more frequently. In dynamic industries, a trend can already be observed that a large part of the master data is no longer up-to-date after 2 years at the latest. This means that it is becoming increasingly costly to keep data in a usable state.
Master Data: Requirements For Automation Using the Example of Machine Learning
The enormous amounts of data generated by digitalisation require new ways of dealing with the related problems. It is becoming increasingly important to use automated processes to filter out the necessary or desired information. Here again, data and master data play an important role. After all, the quality of automated solutions can only be as good as the data quality on which they are based. One example of this is machine learning - a technology that is already being used successfully today in the matching of invoices or the recognition of specific product images. However, in order for machine learning to work as desired, the algorithms usually have to be taught using historical data. Useful patterns can then be derived from this. Here it is important that the existing data - for products, customers or suppliers, for example - is of a correspondingly high quality in order to deliver reliable results.
M2M - Communication Requires Adequate Master Data Quality
Another area is M2M communication. Here, machines acquire the ability to communicate with each other and thus communicate bottlenecks or free capacities to each other or forward information to the responsible personnel. If certain material is missing, the missing parts can be ordered automatically. If a machine announces free capacities for a certain period of time and is equipped with the appropriate tool, the parts in question can be diverted to production. But for the processes to function smoothly, the corresponding data - including master data - must be as complete and correct as possible.
More Complex Master Data Management: International & Cross-Sector Networks
Due to internationalisation and the economic convergence favoured by the internet, old value chains and networks are being reorganised. Partners and suppliers who play an important role in one's own core business are now increasingly located on the other side of the world. Other countries, geographical conditions, laws, languages, cultures - this leads to master data not only increasing in number, but also having more and more country-specific differences. This increases complexity and effort. For example, foreign-language designations increase the effort to ensure correct spelling.
Cloud Computing & Increasing Requirements For Master Data
Another trend concerns new software solutions with the advent of cloud computing. Work, production and business processes have to meet ever higher demands due to increasing requirements. Therefore, it is often necessary to combine on-premise ERP systems with solutions from the public cloud or from several clouds. This in turn requires cooperation and smooth data exchange between the different software. If master data and other data are incomplete, incorrect or duplicated, the problems become all the greater. The more integrated the different ERP solutions and the specific add-ons are, the faster errors spread. Delays occur because data must first be cleansed and corrected, or decisions are made on the basis of incorrect data, which may only become apparent later - but are then all the more costly.
Summary & Outlook
As the discussion has shown, the amount of master data is increasing due to digitalisation and other technological developments. The Internet of Things and the diversity of variants are driving this. Furthermore, it is becoming increasingly important for companies to maintain high standards in data processing. Because these are important in order to cope with the rapidly increasing amount of master and other data due to digitalisation. This is especially true when it comes to automating work and production processes. Because this is the only way companies can maintain their competitiveness today and in the future and meet market requirements.
Here's what happens next ... The second part of this series of articles explains whether the topic of master data management is part of an ERP system or whether it has to be solved in a cross-software manner ...
