This article is based on an online round table " Data was my first love” organised by Pointury on May 5, 2022, during which a group of digital leaders met with advisors from SAP to share experiences about how the availability of large amounts of data has become a driving force for change in many organisations.
In the past few years, a boost in data volumes has allowed companies to gain additional business value. We learned from some examples.
Astara is a joint venture between Alcomotive (formerly Alcadis) and the Spanish Bergé Auto, which sells about 200 thousand vehicles per year in 16 countries. This new company has a business strategy that focuses on digitisation and data in the cloud.
Hans Christiaens, IT Project and Application Manager, talked about the challenges that came with it. It was not just about merging data of two companies but also about uniting classic on-premises (ERP and CRM) data with large data volumes generated by new sources such as
High data volumes exist since several centuries, yet the tools, volumes, diversity of structured and unstructured content has changed. Also the possibility to keep data in the cloud was an important enabler. This created opportunities, yet also costs and other challenges.
This new type of digital transformation triggered by the availability of data should be supported by an organisational change. Setting up a data analytics department makes a lot of sense and many companies have started to do so.
From a technical point of view the key to addressing these challenges turns out to be a strong integration layer consisting of both cloud integration services and on-premise ETL tooling feeding the management dashboards. At Astara a tailor-made “Infohub” solution provides a good example of how data warehousing tools and techniques can be fused with modern cloud based tooling to unlock the potential of all the available data.
At bpost, various ways of elevating the use of data opened up a whole new range of possibilities and solutions. As we know bpost is transforming from a service of delivering letters to delivering packages. Customers have high expectations such as fast delivery and notifications of the delivery.
Kevin Veulemans, Data Strategy & Transformation Manager, explained that new developments in the area of real time tracking of data such as IoT allow operations to react faster. In the past information was only available after scans, which happened only a few times in the entire supply chain. Decisions can now be based on facts rather than on experience and gut feeling. While SW development is still important, the importance of data as a key success factor is growing.
BekaertDeslee is a textile company with 4.200 employees producing the outside of 110.000 mattresses per day. 600 million people sleep on its fabrics every night.
Textile production is dependent on many variables: the selected machine, its age, maintenance and settings, the quality of the fibers, the day of the week, the design of the fabric, outside temperature, humidity, etc. This leads to difficult to predict production waste. It could not be predicted with traditional programming.
Rik Holvoet, CIO at BekaertDeslee, explained how they did a successful project with AI and Machine Learning to predict with a fairly high degree of accuracy how much waste can be expected for a certain knitting pattern with certain yarns on a specific machine. The forecasted amount of waste is passed on to the operators via the production orders, so that they can keep a close eye on the machine and intervene quickly if things deviate from what can be expected. This project has already allowed to reduce waste up to 10%.
Not every question can be solved with AI. If there are too many variables and only a limited set of data is available regression techniques won’t work.
At Carrefour the creation of a new data lake in Google Cloud gave business teams greater access to diverse datasets across products, consumers, and supply chains.
Gregory Pierquin, IT Director - Data, CRM & Loyalty Platforms, explained that with supply chain data becoming much more accessible, Carrefour Belgium can ensure that products, including perishable foods, reach stores and consumers quickly and efficiently. In addition, a better view of consumer trends enables the retailer to create new, personalised experiences for their shoppers, such as targeted promotional offers.
Anonymised, consolidated consumer data is also very useful for producers. This allows e.g. to understand that a certain promotion worked well in a specific region or for a certain public yet not in another region.
When switching to a data driven approach change management is important. Users need to experience the ease of use of new applications and realise the value of data. Also at Carrefour the creation of a data team has been a key for success.
During the meeting data lakes and SAP BTP were often mentioned. Stijn Debever, SAP Analytics Solution Advisor at SAP explained that one of the great advantages of a data lake is that it is data agnostic: you can put all types of data in it, from whatever source.
But agnostic also means: stripped of all application and business context. Suppose you want to analyse the billing history of a certain customer via your data lake. In reality, a billing document in an SAP ERP is composed of dozens of tables with complex internal relationships, which together form a business object: a rich semantic whole that is perfectly understood by accountants. While a data lake is pure heaven for data scientists, it can sometimes be frustrating for business people who rely on real time ERP or CRM data, on the correct business context of this data and on a single version of the truth.