Digital Strategy 2025 - Better Decisions

New tasks for the "data-driven" company: This sounds simple on the surface, yet it is an enormous challenge.

Ever larger, more complex and rapidly changing volumes of data (big data) have to be processed and analyzed. The special task here is mastering the volatility and speed of the markets.

Unpredictability and the increasing speed of change have become the central challenge for management, according to Wolfgang Martin in his study "Business Intelligence in the Digital Enterprise," published in 2015.

This presents management with new tasks:

"Traditional control in the company no longer works. Information is becoming the crucial resource for coping with the complexity and dynamics of the digital world."

Beyond empirical knowledge - on which traditional control was based - decisions must be made quickly today and in the future on the basis of timely information.

"What matters now is having the right information at the right moment that has the right relevance to a decision that needs to be made"

says Martin.

The fact that companies today are "data-driven" - there is no process without data - is the most important driver "for an evolution and, as it were, rediscovery of business intelligence," Martin clarifies.

Business Intelligence (BI) is understood to be all strategies, processes and technologies in which information is derived from data and knowledge critical to success is derived from information, so that decisions are made on the basis of facts that trigger actions for corporate and process control.

The claim of the Business Intelligence concept is therefore to base decisions on facts and to make better decisions.

The problem is that in order to make "better decisions," the facts on which they are based must first be correct. This is no easy task, because the volume of data is already unimaginably large today and its growth is exponential.

Data quality and analysis as BI problems

However, the sheer volume of data is of limited value; only in a meaningful context does it become business-relevant information.

When it is said that companies see the greatest opportunities from Big Data in increasing profitability, followed by a better understanding of the market and optimization of organization and processes (PWC study, May 2014), then merely larger volumes of data are not sufficient for this.

Larger data volumes do not necessarily mean better data quality. However, high data quality is an indispensable prerequisite for drawing the right conclusions from the huge volume of data.

The fact that data quality in particular is not at its best has been proven by numerous studies from recent years, including Steria Mummert Consulting (2013), Barc (2011, 2013, 2014, 2016), Fraunhofer IAO (2013, 2014) and most recently Lünendonk (2016).

In a paper published in January 2015, "Today's data mastery: multiple domains for a single purpose," Aberdeen Group notes that it is the suboptimal analytic environment that both prevents companies from generating tangible value from their data and leaves opportunities for growth and efficiency gains untapped.

Fifty-four percent of the companies surveyed say current inadequate data analysis is the top challenge driving data management activities. In second and third place are too many data silos (49 percent) and poor data quality (47 percent).

In 2017, KPMG, in collaboration with Bitkom Research, examined the status quo and prospects for data analytics in German companies for the third time, following 2015 and 2016. The study "Creating Value with Data" comes to the conclusion that 58 percent of companies assume that the data analyses they use are accurate - but as many as 42 percent have doubts about the accuracy.

This may be due not least to the fact that many companies are struggling with the quality of their data. Around a third of the companies surveyed said that poor data quality was a challenge.

According to KMPG, there are hardly any differences between the industries. Only machinery and plant manufacturers seem to be particularly affected: As many as 51 percent perceive the poor quality of their data as a hurdle.

MDM improves accuracy

Can master data management help? Master data refers to static basic data or reference data on objects relevant to operations, such as products, suppliers, customers and employees.

In another survey on master data management, Aberdeen Group asked 192 companies in September 2014, among other things, whether they had been able to improve the accuracy of their business decisions within a year. V

Of the companies that use MDM, 58 percent say this is the case; in contrast, only 45 percent of companies without MDM achieve this. Better data accuracy in organizations with MDM thus contributes to the increased rate of improvement in the accuracy of business decisions overall.

Master data management not only provides decision makers with improved visibility and more reliable raw data for analysis, it also enhances opportunities for collaboration between internal and external stakeholders, he said:

Stakeholders across different departments, business units or even organizations have access to the same relevant data as they need it.

Multidomain MDM

Perhaps the most defining characteristic of the so-called MDM "leaders," according to the Aberdeen Group, is their ability to manage multiple data domains simultaneously.

Multidomain MDM improves data efficiency in several respects: On the one hand, data accuracy is 8.7 percent better and completeness 11.9 percent better than with non-multidomain MDM. However, the advantages of multidomain MDM are even more evident when it comes to the criteria "time to information" and "accuracy of decisions," which are not least important for effective BI.

When using multidomain MDM, 64 percent see improvements in "time to information," compared with only 35 percent for non-multidomain. When it comes to improvements in the "accuracy of decisions" criterion, multidomain is also clearly ahead of non-multidomain with 69 percent and 48 percent respectively.

Organizations with only one version of the "truth" and a master data record for each key domain spend less time searching for information or confirming the reliability of existing data, and more time on relevant analysis.

Multidomain MDM centralizes the entire master data management. All relevant data from purchasing to sales converge in one central system.

In this way, a "golden record" can be created for customers, products, and suppliers, for example, and connections and correlations between these domains can be identified. You get an all-round view of the master data across all domains.

A multi-domain MDM thus creates the "one truth" for different master data domains across the entire business process.

Whether you are a specialist department or a company management - if you want to gain information from data and mission-critical knowledge from information in order to be able to make decisions on the basis of valid facts, there is no way you can avoid professional data quality and master data management. To avoid this would be to run the risk of not only not making better decisions, but making the wrong ones.