Success Story: Hermes Germany
Hermes transforms into a Data Driven Company
Implementing a strategy together
Hermes Germany has set itself the goal of becoming a data-driven company. In this way, the internationally successful logistics company wants to leverage extensive value creation potential: Via the digitization strategy, solutions are to be developed in the future on a digital image of the entire supply chain in order to bring about a data-driven improvement. This process makes changes calculable, measurable and also sustainable at all times. Such a high level of digitalization cannot be achieved without the use of business intelligence (BI). The necessary basis for successfully shaping this change is a specially tailored BI platform. As an implementation partner, Hermes Germany relies on the proven expertise of PTA IT Consulting in Hamburg.
Hermes Germany GmbH is one of the most successful logistics companies in Germany and Europe. Since its founding in 1972, the consumer logistics company based in Hamburg has become Germany’s largest mail-independent delivery company to private customers in the B2C and C2C sectors. Today, Hermes has more than 1.5 million customer contacts – every day. With around 16,000 Hermes ParcelShops in Germany, the company has the largest nationwide network of acceptance points for private parcel shipping in Europe, which can also be used by mail order companies as alternative delivery addresses. Hermes Germany GmbH is part of the Hermes Group, an international trade and logistics service provider and part of the Otto Group. The range of services offered by the companies operating under the Hermes brand covers the entire retail value chain: sourcing, quality assurance, transport, fulfillment, parcel services, two-man handling and global e-commerce services. In fiscal year 2019/20, Hermes Group increased its total revenue to 3.5 billion euros. Hermes employs 15,563 people worldwide.
Fit for the future
As an internationally successful logistics company operating in a competitive market, it is only natural that the entire corporate strategy should be scrutinized and analyzed at the appropriate time. In the course of this, those responsible came to the conclusion that Hermes Germany’s strategic goal must be to develop into a data-driven company. In doing so, the logistics company is pursuing two thrusts. On the one hand, it aims to make even better use of value creation potential from an external perspective, and on the other, from an internal perspective: With regard to its B2B and B2C customers, the company wants to continue to grow profitably through high customer satisfaction. Indispensable for this are a consistently high level of quality and performance, clear differentiation from the competition and dedicated industry expertise, as well as the successive expansion of sustainable and fair solutions. Looking inwards, the new corporate strategy aims to ensure sound and targeted management capability with the help of a stable IT infrastructure, scalability of performance, and end-to-end digitization of the value chain.
Sophisticated business intelligence strategy at Hermes Germany
In the future, Hermes Germany will develop solutions based on a digital image of the entire supply chain via a dedicated digitization strategy. To do this, the company will set its operational challenges against a digital clone of its business model in order to bring about a data-driven improvement, according to the future approach. Through this process, Hermes Germany makes change predictable, measurable and can make it sustainable. It also makes bold decisions that involve a certain amount of risk clearly calculable.
In addition, management can constantly read the state of the business model, taking into account all the measures taken. In order to implement this digitization strategy and achieve such a high level of digitization, it quickly became clear to all project participants that the targeted use of business intelligence must represent an essential competence throughout the company. A BI platform tailored to this was to form the necessary basis. However, two additional pillars were added: Data Warehouse (DWH) and Data Science (Business Analytics).
The service built on top of this is designed to help the company make data-based decisions through its iterative, systematic investigation methods. “Due to the ever faster pace of technological development, it is becoming increasingly challenging to find the right infrastructures and then keep them up to date with the services for big data and business analytics,” says Sebastian Gießler, Team Leader Business Intelligence Hermes Germany, describing a key challenge of the project.
The central guiding question and strategic project goals
The framework conditions and specifications were set, and so the BI department at Hermes Germany set about designing and implementing the BI infrastructure. The following key question was at the center: How does a BI landscape have to look like to answer every question about the company and its environment on a data basis and is therefore powerful and future-oriented for Hermes Germany?
The BI team formulated three central goals for the three pillars BI platform, data warehouse and data science: On the one hand, the new BI world must connect, process and provide data in near real time. On the other hand, the reporting of the BI landscape must be expanded to include business analytics in order to be able to perform more complex analyses such as forecast models and machine learning in addition to standard evaluations. And it must make self-service BI available to the business departments and enable them, through a modern BI front end, to independently create analyses based on detailed data.
Data flows from heterogeneous sources into a data lake
Hermes Germany brings together diverse information from heterogeneous data sources in a data lake: All accruing internal, external, structured and unstructured data (Internet of things) is collected in real-time and is available unchanged, historized and immediately after generation “near real-time”, regardless of whether it is currently needed.
An important insight: infrastructure, i.e. the BI platform and analytics, belong together. The operation of the platform by the BI department itself reduces dependencies on other departments and thus leads to more speed and requirement orientation.
Architecture stringently implements technological goals
The entire BI architecture is transparently documented. It is comprehensible for the members of the BI department. Hermes Germany can therefore stringently adhere to it within the framework of the projects to be initiated. The architecture developed in this way realizes the technological goals from Hermes Germany’s mission statement:
- It creates a uniform and high-performance database, with all users using the same dimensions and key figures via a single point of truth.
- It views the data warehouse and data science as complementary rather than separate.
Ideally, it makes all data available in near real time.
- A company-wide Business Object Model (BOM) serves as the business view.
- It defines how missing master data will be handled in the future (Oracle APEX).
- It connects all scanners (IoT) and enables Big Data streaming.
- It provides the business departments with a self-service BI by means of clear tableaus.
- It respects data protection and considers data security.
- It enables the storage of original data. So-called hot data is stored for two years in the data warehouse, cold data for eight years in Apache Hadoop.
- It stands out as a changeable, modular and adaptable overall architecture.
Data Lake: The Hybrid Analytical Ecosystem at Hermes Germany
Implement the strategic goal together with PTA
For the implementation, Hermes Germany is starting three projects that result from the formulated goals and – as far as the dependencies allow – run in parallel: In a first project, the project team builds a suitable infrastructure; in a second, a new data warehouse that fits the new infrastructure; and finally, in a third project, the entire data science area is established.
The IT consultants from PTA IT Consulting play a key role in all three projects. They work as part of mixed teams throughout the complex project and demonstrate their expertise in a wide range of tasks. The project results and successes are the result of the symbiosis of the competencies of Hermes Germany’s BI division and PTA.
“PTA has understood our technology needs and is taking a targeted approach to new technologies, subsequently empowering the rest of the organization with jointly tested concepts. This makes the major changes plannable and less risky for us,” says Lars Schröder, Head of BI & Support Solutions Hermes Germany, expressing his satisfaction.
Build suitable infrastructure
In a first project step, the team designed a system architecture for the BI platform. The BI specialists essentially took the following requirements into account: All source systems were connected in real time, batch data integration was switched to real-time data integration, and historized online access to original data was provided in Apache Hadoop, a free framework programmed in Java for scalable, distributed software.
Furthermore, the project team set out to build the central data lake according to the hot/cold principle and to ensure access to both current and older long-term data. As envisaged in the objectives, the departments now also have a modern BI front end that enables convenient BI self-services.
The entire basic construct is equipped with the appropriate data protection measures in accordance with the DSGVO with Protegrity.
Related project information in our project database:
- 5009: Architecture conception and implementation of a BI platform in the sense of a data lake
- 4865: Integration of a Hadoop system into the existing classic Business Intelligence (BI) landscape
- 5113: Evaluation of Snowflake as an Enterprise Cloud Data Warehouse (DWH)
- 5143: Evaluation of the suitability of Exasol as an enterprise data warehouse
- 5119: Monitoring BI Platform
- 5724: Expansion of the monitoring and alerting of a business intelligence platform
- 5389: BI Plattform as a Service
Related project information in our project database:
- 4864: Concept for the implementation of detailed data reporting in the data warehouse
- 5380: Rebuilding a data warehouse according to the Business Object Model
- 5180: KPI Dashboard
- 5381: Monitoring of ETL / ELT processes in the Business Intelligence environment
- 5772: Development of ELT processes for building a data warehouse (DWH) according to the business object model
New data warehouse & data integration become reality
In parallel, the BI department designed and developed a new, generally applicable and integrated enterprise data warehouse based on the Business Object Model (BOM). The BOM represents an abstract view of the essential business processes of Hermes Germany. While in classic data warehouses the data from the source systems is made available in an aggregated form in several data marts, the newly designed data warehouse at Hermes Germany based on the BOM now offers the departments the detailed data in a holistic view, i.e. no aggregates. This enables the departments to generate reports independently in the sense of self-service. The performance of the reports in operational use is now very much appreciated by the departments, which are now variably positioned and able to quickly create both predefined analyses and individual data tableaus via the self-services. In addition, the data warehouse provides (hourly) current company KPIs that can be called up at any time.
Sophisticated analyses with Data Science
The team also pursued another project with the establishment of Data Science. The aim of this is to offer the departments more complex analyses in the sense of business analytics in addition to standard evaluations and reporting in everyday business.
This includes the application of data science methods from the fields of statistics, machine learning and artificial intelligence. Data science is now an established service in the BI area.
Related project information in our project database:
- 5388: Data integration for the purpose of digitising the value chain / supply chain
- 5133: Development of a streaming application for real-time reporting of event data
- 5379: Pricing based on a digital supply chain
- 4960: Reporting for fraud detection using Tableau
- 5746: Generation of mass data for the simulation of a logistics network
- 5774: Development of an AI-based system for forecasting the transit times of parcel deliveries
The following successes were achieved as part of the extensive BI projects implemented by Hermes Germany together with PTA’s IT experts:
- Conversion from batch to real-time data integration.
- From prefabricated reports to self-service (Tableau successfully introduced alongside Business Objects)
- First components successfully migrated from on-premise to the cloud
- Original data historized in Hadoop and still accessible online
- Introduction of data protection (DSGVO) with Protegrity
- Puncture / MVP of the DWH according to the Business Object Model. (Hourly) up-to-date business KPIs can be accessed at any time.
- Data Science is an established service of the BI area
A change project as extensive as the one at Hermes Germany can only succeed if the departments, management and employees follow a jointly developed and coordinated mission statement. This describes the self-image and the basic principles of the BI department. Internally, this mission statement provides orientation and thus acts as a guideline and motivator for the organization as a whole and for its members. Externally, it makes clear what the BI department stands for in the future as part of the corporate strategy.
All employees working in the department, both internal and external, follow this mission statement. It includes technological goals, defines degrees of freedom and responsibility as well as the self-image of a value-oriented service concept, conveys quality awareness, quality assurance and clarity about the necessary data protection. Multifunctional teams have proven their worth as a work organization. In this way, Hermes Germany ensures a value- and requirements-oriented end-to-end delivery capability for data and all processes related to analysis and reporting.
The next project steps are already envisaged
However, Hermes Germany’s transformation into a data-driven company is by no means complete with these three BI projects. The company plans to continue to consistently expand its BI platform. This includes the migration of all components from the Open Telekom Cloud to the Google Cloud.
The project managers are also planning to migrate the data warehouse to the Google Cloud. On-premise operation is to be successively replaced and the first components have already been successfully transferred to the cloud. Hermes Germany is also planning to expand the Business Object Model and implement it in the data warehouse.
And to further reduce the complex IT infrastructure at Hermes Germany and minimize the associated administrative effort and costs, the company will automate the operation of the platform together with the PTA experts.