Digital and Data Platform (abbreviated as DDP) is a reference model for enterprise digital technology architecture summarized by BCG based on the world's leading digital transformation practices.
Enterprise digital transformation has the following characteristics in the application of information technology:
- For users inside and outside the enterprise, with user experience as the center, and constantly innovate digital products;
- Integrate internal and external data of the enterprise and use data analysis to form business insights;
- A large number of business processes have been intelligent and automated;
- Business capabilities are modular, callable services that are rapidly integrated inside and outside the enterprise to support the transformation of platform-based organizations and platform business models;
- Upgrade core systems in a digital environment.
Establish a sound project planning and implementation system in terms of architecture management, business process design, change management, IT delivery model, etc.
Technical architecture requirements for digital transformation
Most of today's enterprise information technology architecture is not agile enough to meet the requirements of digital transformation: the traditional vertical construction of information systems is difficult to modify, integration is chaotic; data is fragmented and locked in various systems, can not be freely used; enterprises lack their own technology development capabilities, can not respond quickly to business needs. Therefore, there is an urgent need for enterprises to build a next-generation information technology architecture.
DDP was proposed by BCG's architecture experts because of this trend. Its main features are:
1. Modularity
Decouple the traditional "megalithic" monolithic architecture of enterprise business application systems to form smaller business application modules that can be independently created, modified, exchanged, and interconnected, and each business application module represents a business capability, called "microservices".
2. Platformization
The way to connect between business application modules is called application application interface (API), which is not only used to connect business capabilities within the enterprise, but also to package and open business capabilities to the outside of the enterprise, so that the enterprise business platform.
3. Hierarchical
In order to form an integrated user interface, business capabilities and services, application systems and data views, and liberate data from traditional systems, DDP is horizontally integrated and divided into infrastructure layer, core business system layer, data layer, and digital application layer.
4. New technology capabilities
DDP has a new generation of technical capabilities, including a combination of software development and testing, so that developers can continuously integrate and deploy for production environments, and establish a DevOps system that improves the efficiency of software application release. In addition, the new generation of architectures exposed to the Internet environment must improve the cybersecurity protection of systems and data.
5, the need for organizational change support
Implementing the DDP requires advancing organizational changes in a range of operating models and technology management. In terms of business operation, it is necessary to improve the ability of platform management and service governance, and fully integrate business and IT organizations. Within IT organizations, it is necessary to develop engineering capabilities, reshape personnel skills, and change the traditional IT supplier outsourcing and cooperation model.
What is DDP?
The DDP model consists of six layers, namely the digital application layer, the data and analytics layer, the business core system layer, the cloud infrastructure layer, the integration and API layer, and the network security layer.
1. Digital application layer
This layer is a digital interface for user use and supports user interaction to provide services. Its responsibilities are first of all to achieve omni-channel access and seamless experience of the user's full journey, such as unified identity authentication, unified process orchestration, etc.; second, based on the microservice framework, to provide fine-grained service components, such as user management, order management, commodity management and other components in the e-commerce application environment; and second, to provide digital general technology components to support the operation of business components, such as face recognition that processes identity, speech recognition for text conversion, natural language processing that supports intelligent interaction, Knowledge graphs that support intelligent decision-making, and more.
2. Data and analysis layer
This layer is also known as a big data platform or data analysis platform, providing data services for business applications, and its main responsibilities are: 1) unified access and integration, storage and processing of various data from different channels (such as the private domain and public domain of digital marketing and the back-end core business system, etc.), including structured and unstructured data; 2) establish a data governance mechanism to ensure the quality of data applications; 3) analysis and services, with the help of various analytical models, algorithms and machine learning capabilities to extract insights from data. Provide intelligent services based on these insights, such as personalized recommendations, precision marketing, pricing decisions, risk warnings, and more. These services are accessed by other applications in the form of APIs.
3. Business core system layer
This layer corresponds to the ever-changing, agile digital application layer described above. We refer to the information systems that support the more stable, infrequent changes in the enterprise, and process internal control information records and business processes as "steady-state systems", which are not necessary or difficult to decouple in the short term into microservices, and such systems belong to the core business system layer, that is, the "legacy systems" inherited from the traditional enterprise information age. The above two layers are conceptually close to the "business middle office" and "data middle platform" that are popular in China, so this layer can be considered as the "business background", such as the familiar enterprise-level solutions - ERP systems, HR systems, CRM systems, and various business execution systems (such as the MES system of production enterprises, the core trading system of bank securities, the insurance industry's insurance policy and claim system, etc.).
4. Cloud infrastructure layer
Infrastructure is a variety of computing, storage, and network resources that support the operation of enterprise information systems. Traditional infrastructure takes the form of a physical computer room or data center, while the current mainstream form is provided by cloud providers in the form of cloud services. The simplest form of enterprise cloud migration is the virtual machine running on the cloud by the application system, while the purer way to go to the cloud is that the development and management of the application system adopts the cloud platform. The cloud increases the elasticity of enterprise spending on infrastructure, and enterprise application system deployment can be selected from a variety of modes such as private cloud, hybrid cloud, and public cloud.
5. Integration and API layer
The various services, systems and tiers of the enterprise need to be interconnected, and integration technology and APIs act as glues. Internally, the information flow integration between each level and each component within each layer is realized, and the system information flow integration between the external implementation and each ecological partner is realized. Integrations include traditional message integration (MQ) and enterprise service integration (ESB), as well as API integration in the digital age.
6. Network security layer
In the digital age, security risks have penetrated into all levels, such as identity authentication, access control, application security, data security, host security, network security and physical security, etc., and a sound security mechanism needs to be established at all levels.
Modernization of core systems
In the traditional process of enterprise informatization, a number of core business systems that support internal business processes, record control and compliance-related information have been formed, the most representative of which is the well-known ERP system that realizes cross-functional integration. Due to the limitations of technical architecture, such core systems have a large number of manual processing in business operations, user operation interface is not friendly, transaction operations and data analysis are separated and real-time analysis and decision support cannot be realized, secondary development costs are high, and integration is difficult.
The implementation of the DDP architecture can change these deficiencies of traditional core systems, and effectively promote the shift from informationization with process control as the core to digitalization with user experience as the core. The key point of the transformation of the traditional core system to the next generation architecture is "kernel replacement, decoupling, and cloud migration".
Point 1
Swap the core
The technology of the traditional core system is being upgraded, for example, the most mainstream ERP system on the market, SAP, has launched a version that adopts a new generation of database and development technology - S/4 HANA, and many companies using the old version of SAP ERP are upgrading to S/4 HANA to build a new generation of digital cores; steady-state business recommendations remain in the core system to form a streamlined core.
Point 2
decoupling
For services that support continuous changes in the user's front end, or business capabilities shared across systems, they need to be separated from the core system in a modular manner, a process called "decoupling", so that the enterprise application architecture forms a "simple core, loose coupling" form.
Point 3
Go to the cloud
The development environment, development tools and runtime environment of the enterprise application system are transferred from the ERP system to the cloud platform, and all services are accessed through an integrated dispatch center, and application integration, data analysis and AI expansion are realized on the platform.
The following diagram shows BCG Platinion's approach to planning and implementing a "next-generation ERP" based on the DDP model:
Democratization of data
In traditional enterprise information technology applications, due to technical and architectural limitations, data is locked in various information islands. In the context of digital transformation, data is liberated from legacy systems and becomes an enterprise-level asset, and each team or department in the enterprise can independently access authorized data as needed, completely releasing the business value of data assets.
DDP's data layer integrates all the internal and external data of the enterprise and "democratizes" the right to use. Enterprise segmented data sources, whether they exist in internal core systems, real-time trading systems, or external data sources, such as social media, meteorological data and other big data, can be centralized and integrated into the "data lake" through automated methods, pre-processing data (such as data virtualization, data mart, etc.) for actual business applications or analysis scenarios, and then quickly, easily and intelligently generate the required data services. The data layer can also access the streaming data generated continuously during service events to support real-time service response.
Data availability and data quality determine the value of data applications, and valuable data is an asset of an enterprise. Therefore, enterprises need to establish a good data management and data governance system. DDP's data layer also includes master data management capabilities and data governance tools.
API Management
APIs play a key role in digital transformation and innovation by supporting platform and ecosystem business models as well as agile business processes. Microservices and APIs are the two technical foundations for organizations to build platforms and develop ecosystems.
With the implementation of the DDP architecture, the number of APIs within the organization, not only in the IT department is growing very rapidly, but also within the business line, the demand for public and private APIs and API-led connectivity is growing, driving the need for API management solutions in the organization.
DDP needs to have some API management capabilities like this:
1. API Gateway
The component that precedes the API is a single entry point for the defined backend API and microservices, both internal and external. It acts as a protector, reinforcing security and ensuring scalability and high availability. API Gateway receives all API requests, determines which services are needed, and combines them into a unified, seamless user experience.
2. Developer Portal
It is a bridge between APIs and their application stakeholders to locate, promote, and govern the developer community that uses APIs. It also provides other useful features such as runtime management, API analysis, and gives an estimate of the value of the API.
3. API life cycle
Proven tools can support the development lifecycle of software and APIs, delivering in a timely and high-quality manner by facilitating teamwork and task automation. The process of the API lifecycle can be mapped to the various stages of the DevOps pipeline.
Implementation considerations
DDP is the technical grasp of enterprise digital transformation, and the implementation of DDP is not only a technical issue, but also requires supporting organizational, process and personnel skills changes. BCG has identified six key success factors in several DDP planning and implementation projects around the world:
Factor one
Governance mechanisms for digital transformation
Digital transformation is not only the responsibility of the IT department, the agile method of business and IT personnel fully integrated to create digital products is the basic way of working to implement DDP, and large-scale agile promotion within the organization requires the establishment of a new business/IT governance model.
Factor two
With the power to unlock the value of data
to initiate an architectural transformation
Today, most enterprise information technology applications are not a blank sheet, and implementing DPP requires finding an immediate entry point. We've found that freeing data from legacy systems is often the most pressing need for enterprises, and data layer integration is DDP's first choice for quick wins.
Factor three
Business scenarios drive platform construction
DDP serves user scenarios, and enterprises need to design business scenarios according to user journeys, such as digital supply chain, intelligent manufacturing, real-time intelligent recommendation, etc., and put forward requirements for the microservice capabilities of digital platforms from business scenarios.
Factor four
Break down organizational barriers
DDP needs to establish a product management and operation system for platforms and digital products, breaking down organizational barriers within the enterprise according to the requirements of end-to-end business processes or seamlessly reconstructing user journeys.
Factor five
Develop new skills for people
Change the way IT operates
DDP updates the technology stack of enterprise information technology, and the platform puts forward higher requirements for enterprise technology development capabilities, and DDP requires technical roles such as enterprise architects, technology and data architects, data scientists, DevOps engineers, full stack engineers and so on. Businesses should acquire these competencies through retraining or even recruiting new people. At the same time, the implementation of DDP also requires enterprises to comprehensively evaluate the customization of packaged software, IT outsourcing model, code management, open source and other strategies.
Factor six
Define the measure of success
Advancing digital transformation requires a small-step agile approach, and enterprises need to establish a set of metrics to evaluate the progress, maturity, and business value of DDP architecture transformation to ensure the effectiveness of agile iterations.
In order to help enterprises achieve DDP architecture transformation, BCG has established a multidisciplinary team, including industry experts, digital strategy experts, architecture experts, data scientists, network security experts, agile coaches, etc., to help enterprises solve the problems encountered in all aspects of architecture transformation.
Author: Guo Chen, Managing Director of Platinion, Boston Consulting Group (BCG), member of BCG's global DDP Leadership Team, Head of Enterprise Solutions Asia Pacific; responsible for architecture planning and digital product design for enterprise digital transformation.
Jing Jiang, Managing Director, Platinion IT Architecture, Boston Consulting Group (BCG).
Yu Chuan, Managing Director, Platinion IT Architecture, Boston Consulting Group (BCG).
Daniel Jansen Chief Architect, Platinion, Boston Consulting Group (BCG).
Susaijie, Senior Architect, Platinion, Boston Consulting Group (BCG).