2.7 Data Analytics and Integration

Organisations create a vast amount of valuable data, which should be utilised by different businesses and functions. Data analytics and integration maximise the business value of data in organisations by:

  • Providing data analytics and reporting services
  • Making good quality data accessible in a governed, secure, and ethical fashion
  • Providing integration platforms and RPA (Robotic Process Automation) services ensuring data flows efficiently and controlled through the enterprise and value chain
  • Encouraging automated data entry and gathering
  • Keeping master data consistent.


Data analytics

Data analytics (data mining) provides valuable business information based on analysis through various dimensions using data from different sources. Typical sources of data are business solutions, data warehouses, publicly available data sources through APIs, and operational data from sensors.

Data analytics activities include inspecting, cleansing, transforming, and modelling data with the objective of discovering useful information, drafting conclusions, and supporting decision-making for the business.

Figure 2.7.1 Overview of data analytics


Value is created by enabling data use for:

  • Creating new insights to support decision-making
  • Providing input to machine learning and artificial intelligence
  • Developing algorithms for predictive analysis, understanding customer behaviour, and operational and asset optimisation.

Predictive analytics uses many techniques, for example, data mining, statistics, modelling, machine learning and artificial intelligence. It also analyses current data to make predictions about the future.

Insights in customer behaviour can be gained by testing simultaneously two alternatives (A/B testing). Data is collected through experimentation with different customer groups and correlated with different service or product offerings.

Real-time availability of data and IoT-based sensor data supports optimisation of operational processes, services and assets.


Data ownership and governance

Assigning business ownership to selected data domains and data sets is a good way to ensure that the business-relevant data is kept consistent, of good quality and up to date. Data ownership is a business role and should be recognised for the potential business value and for the potential business risks it represents.

Whereas general accounting principles do not demand full transparency on the intangible value of an organisation’s data assets, some organisations have started to report the value of their customer data. Determining an absolute monetary value on data assets can be difficult, but the relative value change can be measured by tracking the actual use cases that consume or create new data assets.

Data governance relies on implementing data-ownership models and assigning clear accountability with effective operational roles. Using data requires good governance practices to ensure compliance with regulations, such as GDPR and data security.

The Business Technology Data Officer (BTDO) has overall ownership and accountability for data usage; a role that can be deputised to different data or business domains.

The operational data lead and expert roles can be tailored as needed for a centre of excellence type of service organisations for example. This type of organisation develops new data and analytics capabilities and provides them as services to the business.


Figure 2.7.2 Levels of data management



Business processes span across different business solutions and data sources. Therefore, data flows and integrations are an essential part of business technology management. Without integrations, the data stays in silos serving needs of a limited group only. When data is taken out of silos and combined with data from other sources, it allows new views especially from a reporting and analysis points of view. Data flows have also an essential role in automation between multiple systems inside or outside the organisation.

There are several alternatives to implement integrations:

  • Manual integration is done as integral part of a person’s job or as their main task. Every organisation has numerous low-volume integrations done by humans. High-volume manual integrations will be automated or are typically done by low-cost labour
  • Software robotics automation (or Robotic Process Automation, RPA) is manual integration done by a software robot simulating a human user. A lot of manual work can be replaced by software robots without making any changes to business solutions. This extends the lifecycle and cost-effectiveness of legacy solutions
  • Point-to-point integration is an integration between two business solutions with defined protocol and data structure. Implementing multiple point-to-point integrations can be costly and difficult to maintain but enables high volumes
  • Enterprise integration traditionally covers Enterprise Service Bus (ESB) and Extract/Transform/Load (ETL) functionalities and is implemented with an integration platform managing data structures and multiple integrations. Changes in data flows and structures are easy to manage and possible issues in data transitions can be monitored
  • Application Programming Interface (API) integrations enable integrations with one system initiating processes in another system and enabling thus a full process integration.

Complexity of integrations varies a lot and is often underestimated because integrations usually contain other aspects than just the technical integration. For example, security, access rights and error handling requirements may cause extra work to integrations. In many cases, organisations can reduce the number of required integrations by having extensive application platforms. When data and process consistency is guaranteed by the platform, the integrations become seamless.