Alibaba, the most valued retailer, has no warehouses.
These are the most prominent but by far not the only examples for the disruptive power of cloud computing. In January 2016 VentureBeat counted 229 so called unicorns – start‐up companies valued at over $1 billion. The largest include platform companies like Uber and Airbnb, Palantir, Snapchat, Pinterest, Dropbox or Spotify – all of them operating a business model directly based on cloud computing.
61.3 Becoming a Digital Business Starts with Cloud Services
Cloud computing technology and a flexible consumption‐based price structure associated with off‐premises hybrid, private, or public cloud compute models have created the ability to deliver new offerings to market, which were simply not achievable in the past.
But cloud technologies not only allow the creation of new and highly disruptive business models, they also enable the transformation of existing businesses. By implementing a strategy to capitalize on the cloud, companies can stop just running business and start making it thrive.
Becoming a digital business starts with cloud services and cloud solutions gain rapidly in importance on corporate management agendas. According to a joint study by BITKOM and KPMG, the attitude of German companies towards cloud computing has improved significantly in recent years. Today 54% of all German enterprises use cloud computing solutions while an additional 18% are planning or considering their introduction in the coming years.1 Smaller businesses that were hesitating in the past are now massively catching up (see Fig. 61.2).
Fig. 61.2Application of Cloud‐Computing
Assessing the overall cloud computing market, the Experton Group forecasts growth rates in the high double‐digit percentage range for the year ahead. For instance, in 2016 growth is forecast at 35% leading to an overall market value of nearly EUR 12 billion.
While security concerns are still the major obstacle to the proliferation of the cloud in Germany, new offerings designed to meet the special needs of German customers will further foster the demand for cloud solutions in the near future (see Fig. 61.3).
Fig. 61.3Requirements of Cloud Computing Customers in Germany
61.4 Offering Customers More Flexibility and Choice
As for German customers it is a major requirement that their data is hosted exclusively in Germany, several cloud providers have started to open new datacenters in Germany. Microsoft is the first public cloud company to deploy a data access model where a German Data Trustee controls access to customer data in accordance with German law. Operations or other tasks that require access to customer data or infrastructure will be performed or supervised by the Data Trustee, T‐Systems International. As Microsoft does not have access to customer data without prior approval by the customer or the Data Trustee, it is unable to respond if a law enforcement or other third party request for customer data occurs. The Data Trustee will not disclose customer data except where required by German law. Data between the German datacenters is exchanged through a private network to guarantee that data resides exclusively in Germany. Additionally, Microsoft maintains redundant sites in Germany to ensure business continuity and customers are at any time able to view how and where data is processed. The new solution offers customers increased flexibility and choice and has the potential to spur local innovation and growth.
61.5 Drawing Insights from (Big) Data
Cloud computing offers scalable data collection, processing, and analysis capabilities flexible enough to adapt to the needs of potentially every business. Above all, the power of the cloud to harness, store, and draw insights from (big) data is a game changer. Big data – a collection of datasets so large and complex that they become difficult to process using on‐premises database management and processing applications – needs a flexible, scalable compute model that evolves as the business evolves. Big data has to be contextual and, through its very nature, combined with many other assets, sources, and datasets.
Only cloud solutions give businesses the ability to process significant amounts of data, whether it’s latent or in real time; store that data; and then apply rules and structure to it for consumption. The cloud also enables even more data to be unlocked by enabling businesses to pull data in from different sources and across different line‐of‐business assets and devices.
When using cloud‐based solutions for storage and analysis, companies can combine data from multiple sources without worrying about capacity constraints or the significant costs that might result from building out on‐premises infrastructure and automate—through filters, rules, triggers, or other means—the intelligent processing of that data.
61.6 Responding More Quickly to Competition and Changing Market Conditions
In the cloud companies can unlock new value by making the best use of both existing and new data sources. Deep mining of data from disparate sources creates new insights and even predicts future outcomes. Analyzing data that has been acquired over a long period of time is helpful to find patterns and correlations to uncover trends that offer new insights about how products are used or how customers behave under certain conditions. With machine learning capabilities from the cloud, companies can apply historical data to a new problem by creating a model and using it to successfully predict future behavior or trends. Data insights from the cloud help to respond more quickly to competition, supply chain changes, customer demand, and changing market conditions.
The Experton Group expects the German big data market to grow from EUR 1.4 billion in 2015 to almost EUR 3.8 billion in 2020. At present, big data technology in Germany is largely driven by the internet, e‐commerce, and advertising sectors. However, thanks to its competitiveness and export orientation, the German economy is expected to quickly adapt to the needs for optimized production, logistics, and sales process to become an international “big data champion,” according to BITKOM2.
61.7 Connecting Things to People and to Services
Cloud computing not only helps to run business more efficiently by simplifying and automating operations and processes with data and intelligence. It also has the power to create new customer experiences and to support the invention of new product offerings and business models. Last but not least cloud computing is also about engaging and empowering employees by enhancing collaboration and supporting new mobile workstyles.
To take full advantage of all these possibilities, bringing data into the cloud is not enough. We need to build systems of intelligence around customers, employees, operations, products and business models. Today, leading companies are exploring the correlation of connecting things to people and to services, creating a new form of intelligence synergy.
61.8 Starting with Small Changes for Big Impact
That might sound complex. But it is a long‐term process that can start with simple steps, marking small changes for big impact. In the beginning it is important to focus on the areas of business that provide quick return. Digital transformation should start with identifying the one process, product line, or location that might matter the most, for example: Connecting robots on the factory floor with back‐end systems to create a production line with more continuous uptime.
Sharing diagnostic images from a CT‐scan machine in near real time with radiologists at another medical facility and the family doctor to improve patient care.
Connecting one handheld device to the inventory system to get real‐time customer service on the sales floor.
Comparing results from different store locations to identify the most successful services and roll them out nationwide.
Connecting point‐of‐sale scanners on a retail floor to warehouse systems and analytics software at headquarters to increase efficiency in inventory.
Adding expiration dates to the data set for pharmacy inventory to save substantial amounts of money in wasted medications.
61.9 Creating a More Prosperous Economical World
In the middle of the next Industrial Revolution, companies need to ensure that they stay relevant in their unique competitive markets. They need to be able to engage with their customers in new ways to drive loyalty and growth, and use digital insights to engage with them more personally. They need to empower their employees to be more effective across all aspects of their jobs, they need to optimize their operations to drive efficiency and they need to enable new products and business models.
The proliferation of the cloud and connected systems and devices allows industries to be more open to collaborate and to share practices and information from production to delivery of a product or a service. With the advent of machine learning and new intelligence tools with enterprise security on the end point, these budding ecosystems are ready to become more intelligent than ever, thus promoting greater margins for all players and in general, creating a prosperous economical world with accelerating GDP.
Footnotes
1 https://www.bitkom.org/Presse/Anhaenge-an-PIs/2016/Mai/Bitkom-KPMG-Charts-PK-Cloud-Monitor-12-05-2016.pdf.
2 https://www.gtai.de/GTAI/Content/EN/Invest/_SharedDocs/Downloads/GTAI/Fact-sheets/Business-services-ict/fact-sheet-software-industry-en.pdf?v=3.
© Springer-Verlag GmbH Germany 2018
Claudia Linnhoff-Popien, Ralf Schneider and Michael Zaddach (eds.)Digital Marketplaces Unleashedhttps://doi.org/10.1007/978-3-662-49275-8_62
62. Data Virtualization: a Standardized Front Door to Company-Wide Data Opens the Way for (Digital) Business Success
Christian Kurze1 , Michael Schopp1 and Paul Moxon1
(1)Denodo Technologies, Munich, Germany
Christian Kurze (Corresponding author)
Email: [email protected]
Michael Schopp
Email: [email protected]
Paul Moxon
Email: [email protected]
62.1 The Need to Go Digital
In recent years, power has shifted from the companies to the customer. With a few clicks, customers can check competing offers and easily switch from one vendor to another. The customer, in short, is king, and expects to be treated like one. Today, to attract and retain customers, businesses need to transform from being product‐centric to being customer‐centric, and strive for business continuity in all processes. Such a transformation is not easy, despite the fact that the world’s information, much of which is made up of customer and machine data, is doubling every 1.5 years, and is expected to double every day by 2050. Buckminster Fuller has first stated this exponential growth in his book “Critical Path” back in 1982 [1] and proves to be right nowadays with the advent of Big Data and the Internet of Things (IoT).
Such a transformation requires businesses to be able to quickly access massive databases of customer data and immediately respond to customers’ needs, even when their needs change in a few minutes, such as after a purchase or a similar compelling event. Such a transformation also requires an optimized supply chain and a completely reimagined value chain. It also requires an open dialog between management, business departments, and IT about streamlined integration across business functions, with an increased focus on project outcomes, better decision‐making, efficient operations, and an orientation towards developing new products and services that meet changing customer needs.
The biggest impediment to digital transformation continues to be massive quantities of heterogeneous data, since it is located in separate silos, requiring costly, time‐consuming integration, or multiple, manual steps to process a simple query across the entire data set. Nearly all initiatives driven by digitization require rapid data integration in one form or another, including the deployment of new APIs, big data and/or cloud sources, mobile solutions, SaaS offerings, and interfaces with the Internet of Things.
The model of the monolithic enterprise data warehouse (or data mega store) fades away, and data heterogeneity has come to be accepted as the new normal. Hadoop, Cloud, NoSQL and other new sources appeared rapidly during the last years. This new world of distributed and diverse data needed by many apps and users is real, and it will not go away. Such a world demands that businesses develop a fast data strategy; otherwise, businesses will simply not be able to leverage the wealth of data that is already in their hands. As Forrester [2] says, “Business stakeholders at the executive and line‐of‐business level need data faster to keep up with customers, competitors, and partners.” Well‐known numbers underpin the need for timely data availability: Fixing a product after delivery costs 10 to 30 times more than during its construction or production process. Retaining an annoyed customer may cost $100 in discounts, agent calls, and process costs, whereas fixing the problem earlier could cost as little as $5.
Data virtualization enables the use of agile, real‐time, self‐service data technologies that deliver data to business users in real or near‐real time, to effect faster outcomes.
62.2 Modern Strategies and Architectures
Many companies begin their transformation by building data labs and analytics, and establishing data science teams within the business departments. But the biggest challenge is how to provide these teams with all the data while still maintaining compliance.
Data silos, as we mentioned above, hinder the flexible access and shared usage of data across the organization. Ideally, all analytics, reports, processes, and applications (web, mobile, desktop) should see exactly the same customer, product, and partner data.
To meet this ideal, a variety of vendors proposes technology solutions and architectures. Gartner categorizes them into three integration and semantic layer alternatives [3]: 1. Applications and business intelligence tools as the data integration or semantic layer: This approach delegates data integration to end‐user tools and applications, but results in a duplication of effort, since it is necessary to perform integration multiple times in different tools, so changes in the back‐end would require reengineering on the front‐end. In addition, the primary focus of end‐user tools is not to function as integration middleware, but as user‐friendly applications.
2. Enterprise data warehouses as the data integration or semantic layer: In this scenario, infrastructure vendors provide access to data not stored in the data warehouse in a pure query federation mode, often coupled with the traditional replication of data into the data warehouse. The data warehouse, as the integration and semantic layer, remains the “center of the data universe.” Although this approach appears attractive to organizations that are already heavily invested in enterprise data warehouses, it does not address the big picture. What happens when there are more than one enterprise data warehouses (often based on different technologies)? Not all data can be replicated into, or accessed via, the data warehouse, and project and storage costs would increase.
3. Data virtualization as the data integration or semantic layer: Moving data integration and the semantic layer into an independent data virtualization platform leverages the native capabilities of such an abstraction layer to access data across multiple heterogeneous data sources. Recommended by Gartner, this approach provides business‐oriented models for the underlying data via a logical‐to‐physical mapping. Moreover, the virtual layer enforces common, consistent security and governance policies. Advanced data virtualization solutions provide – in addition to the commonly accepted relational access to data – native support for complex data structures commonly found in recent big data technologies, which are effective for providing timely data to any data consumers.
We argue that an optimal fast data strategy is built around data virtualization, since it establishes an intelligent abstraction layer above the heterogeneous sources, which acts as a unified data access layer across the en
tire enterprise. Data virtualization makes it possible to combine any kind of data, such as big data and streams from the IoT, with existing data assets like customers and products.
62.2.1 How Can Data Virtualization Transform a Business?
Let us take a closer look at the transformation to customer‐centricity, effected by leveraging data virtualization.
To support cross‐ and up‐sell initiatives, sales teams need complete, updated information about the customer as well as related information about products, channels, and warranties. Marketing, support, and executive teams all need access to the same information for their different purposes.
The typical IT architecture presents a challenge, since it is often the result of more than 20 years of development and, as we mentioned above, is often characterized by siloed data stored across many disparate systems. In these architectures, each department accesses different systems in different, manual ways and IT responds with multiple point‐to‐point data integrations and even more data silos for each application. As a result, business users do not get answers in time to complete their business. As Forrester [2] says, “data bottlenecks create business bottlenecks.”
Digital Marketplaces Unleashed Page 94
