The first step is to identify the business requirements to eventually design or acquire technology capable of delivering those needs. The selection of technology should be a careful process that analyses the many trade-offs. In the beginning stages, it is important to prove the value of analytics so it might be attractive to use a Software as a Service (SaaS) model in the cloud, where the hosting and powering of a system are taken care of by an external provider. This usually requires lower upfront costs and faster bootstrapping. Many technologies are available this way. You can also take on more burden yourself by using an Infrastructure as a Service (IaaS) model. With IaaS the infrastructure, such as servers, are provided on which you install the applications you need. The trade-off is between speed, flexibility, and costs. SaaS is often the easiest to get started with, but may no longer fit your needs over time and costs more. The technology in this stage focuses on the starting steps of an analytical platform where data is collected and made available in reports and dashboard for management. These dashboards provide an overview of historical and current performance, and leave generating insight to the ones viewing the reports.
Identify business requirements for a basic analytics platform.
Decide if you want to build or buy software and infrastructure.
Where in the Reporting stage the infrastructure was largely ad hoc, a professional analytics platform will be built during this stage. The implementation details will largely depend on the organization's requirements. There has to be a standardization of technology, and a modular platform that is capable of easily integrating new features and components. All data will be stored, in a standardized format, in a data warehouse or data mart, and is easily accessible for services. Invest in tools that help you analyze and mine the available data. A lot of technology is being developed to help the data scientist in making the available data more explorable and increasing their efficiency.
Professionalize the data infrastructure.
Standardize technology, integrations, and warehousing.
Acquire technology that helps data scientists analyze data.
The technological challenge is to integrate these advanced machine learning algorithms in production systems. There is a massive amount of data and calculation necessary to train these models. Timing is a big challenge: New data sometimes has to lead to insight only moments later. The infrastructure has to be able to quickly transport data. Batch processing has to make way for streaming data. From a pull structure, i.e. the data is queried when needed, to a push structure: When new data is generated, an event, it is pushed and processed immediately. Event-driven architectures are build to react to and handle incoming events. Personalization is a big part of optimizing the customer experience. Automated integration of analytics allows us to adjust the customer journey for every unique individual. This is unfeasible to do by hand for everyone. Machine learning algorithms can prescribe the best journey for every customer. Data collection about the customer is needed to serve as input to the analytical model.
Invest in technology that can automatically process new emerging data in near real-time.
Invest in technology to automatically integrate analytical insight into production systems.
Technological developments in this phase allow business users to do their own form of analytics. You provide the tools and they provide the domain knowledge. Self-service business intelligence is the concept of granting tools to business users to build basic analytics functionality themselves. They can build their own reports or analyze data. The caveat is that these users may not be well versed with data and statistics. It is easy to draw the wrong conclusions when you misinterpret statistics. We are also increasingly building in analytical features into our products and services. Features that add value to the customer by providing them with analytical insight based on the data the product or service collects, e.g. a smart thermometer that learns your preferences.
Focus on a modular and scalable infrastructure that can rapidly support new opportunities
Augment products with integrated analytical features (smart products)
Create more opportunities to use self-service business intelligence
Technology is continuously evolving. When you compete on analytics, you are competing on technology. Competitors that are able to better predict customer behavior are more likely to attract and retain customers, while also building a strong relationship that is mutually beneficial. Building and maintaining analytical systems takes a lot of work. We are seeing more software platforms that accelerate the process of building models, allowing data scientists to rapidly churn out new models and reduce the time to market \cite{analytics2018predictive}. With automation happening at every step of the data science process, the data scientist can focus on the most important task: asking the right questions.
Invest in software platforms that accelerate the process of building analytical models
Focus on asking the right questions