Data and its analysis are among the top priorities of companies in the digital age: According to the “State of the CIO 2020” survey by our US colleagues at CIO.com, 37 percent of the IT decision-makers surveyed are planning the largest investments in this field – ahead of time the areas of IT security and risk management.

Photo: fizkes – shutterstock.com
With more investment in data analytics, however, the pressure increases to deliver good results. From an analyst’s point of view, however, these often fail to materialize: “For many CIOs and business decision-makers, getting analytics initiatives across the finish line is a challenge,” says Brad Fisher, partner and US chief for data and analytics KPMG.
In this article, we have summarized the areas in which data analysis projects are particularly difficult.
Nobody really disputes that data has become a critical factor for success. As part of a Gartner survey, 86 percent of IT decision-makers indicated that their companies would give away a competitive advantage if data were not used effectively. However, the same survey also shows that more than half of these companies do not have a data governance framework or their own budget for data management. Failure to do these basic things can wreak havoc on data analytics ambitions.
Without a fully implemented data governance program, there can be no adequate standards for data hygiene. In this case, it becomes more than difficult for companies to access existing data or to integrate it, since it is locked in departmental silos. Such companies may not even know what data would theoretically be available to achieve certain project goals.
In many places there is also a lack of the basic technologies that would be necessary to achieve ambitious data analytics goals. Instead, tools are often used that are hip, but do not match the individual needs of the company at all. In some cases, it “works” the other way around: There, tools are kept that you have got used to – even if they are not innovations.
The basic problem here is that there was no right strategy from the start. If IT decision-makers implement a data initiative along a clear strategy and – even better – a center of excellence has been set up for it, there is a high probability that the basic success factors are in place: data governance, defined responsibilities, infrastructure, training programs, strategic goals and an appropriate leadership culture.
Conversely, this does not mean that companies should view analytics as a monolithic project: Some CIOs go straight to the brim, create data lakes and implement cost-intensive infrastructure components – only to subsequently find that the technology is used far below its capabilities or even is completely ignored.
Implementing solutions that focus on a specific problem is a better way to communicate the added value of technology to users. In most cases, the specialist departments do not have the time to wait for the results of large-scale transformation projects: they want to see results, even if the achievement of goals is not 100 percent right from the start. A success rate of 60 to 70 percent can be built on to gradually optimize. This procedure also has the great advantage that quick success favors further investments.
It is also recommended for companies to base their analytics investments on business cases – not on the emergence of new technologies. An incremental expansion of data initiatives with new, advanced tools should also enable users to grow into more complex scenarios and solve corresponding problems.
Despite massive investments in their data programs, many decision-makers find it difficult to reap the benefits of their work, as the “Big Data and AI Executive Survey 2020” by NewVantage Partners shows: 74 percent of the companies surveyed are still struggling with it, the employees and the organization to become familiar with big data technology. According to experts, this is partly due to the fact that users’ needs are too often not recognized and taken into account.
If, for example, there is a lack of a cross-departmental, holistic data strategy, there are serious inefficiencies. Sometimes user groups are formed that largely have to do without support. At the other extreme are companies that, without exception, centralize everything that makes rapid scaling impossible and prevents the full potential of a data analytics project from being realized.
“A healthy mix of centralization and decentralization is needed, with which the balance slowly shifts. It is advisable to start with a central approach,” recommends Roy Singh, partner at the management and management consultancy Bain & Co.
- Steve Oluborode, Tableau Software
Data is the new oil. A glance at the ranking of the world’s most valuable companies shows that this is no longer a forecast, but rather a reality. The top 3 all achieve their added value by monetizing data. - Carol Stockinger, IDG
The data analyst’s job is anything but new. But it has changed a lot in recent years. In the past, the aim was to prevent duplicates and to maintain data quality and security overall, but today the focus is on the creation of usability as a whole. Do I understand my data? How can I merge, classify, analyze? These are the questions we are facing today. - Michael Koch, Lufthansa Industry Solutions
The essence of the Germans is to want to understand everything in detail. With the gigantic amount of data that is generated in companies, this is simply no longer possible today. Perhaps this is the explanation for why everything is moving a bit slower in this country. - Andreas Laux, Datavard
We have more technological options available than ever before. But realizing the better use of data is a cultural task that customers and service providers can only solve together. It is important to keep pointing out to people how important data is for the improvement of business processes and the emergence of new services. If I can credibly illustrate the resulting added value, then the willingness to “share” increases. - Peter Jung, board
Business is becoming ever more dynamic. Structures, business models and ownership are constantly changing. We have to react to this dynamic with flexible data management: Every day there is a new “data treasure” to be collected and used, that means to gain and provide decision-relevant insights from the data. - Andreas Heißler, Uniserv
The Federal Government’s initiative for its own data strategy sounds less like a “real” strategy. The problem is the great uncertainty within companies about what they are legally allowed to do and what not. Just the parallel existence of various contradicting laws and regulations creates a lack of transparency that inhibits progress. What is right today can be wrong tomorrow. This is a pro
blem for small and medium-sized companies in particular: to establish a functioning data management system, I have to spend money in my hand and that is easier for large corporations to manage. Smaller companies cannot “just try it out”, they need planning security. - Oliver Schröder, Informatica
In Germany, we still lack the speed to adapt business models. The platform economy in the USA already has clear competitive advantages in organizational terms. An obvious indicator is the organizational importance of IT. Many companies still have separate IT departments, and the CIO reports to the CFO. All of this would no longer be necessary in an agile structure in which IT and business ideally merge. - Peter Küssner, Cubeware
The too cautious use of data in the development of new business models is not a technical and not an organizational problem, but simply: a German one!
Nevertheless, IT decision-makers need more than a holistic approach to analytics success that is in line with the strategic goals. A change in corporate culture is also essential. Finally, users have to use data-driven insights in real time and internalize the work with these technologies or see them as a new standard.
According to the NewVantage Partners survey mentioned above, only 38 percent of those surveyed see themselves as “data-driven enterprise” and only 27 percent are convinced that they have established a data culture in their company. There is a lack of well-trained users and the right processes – 91 percent of those surveyed see these as the main obstacles on the way to realizing these goals.
According to KPMG expert Fisher, CIOs and business decision-makers should integrate their data analytics programs into the process world in such a way that users perceive data-based decisions as business as usual. “In most cases, users don’t care what the data sources look like or how cool the underlying data science is. They just want to get the information they need to get their job done Technology feels like an app – a requirement that every CIO should understand. “
This article is based on an article from our US sister publication CIO.com.