Ferris wheel
Ferris wheel (source: Skeeze via Pixabay)

More than 10 years after big data emerged as a new technology paradigm, it is finally in a mature state and its business value throughout most industry sectors is established by a significant number of use cases.

A couple of years ago, the discussion was still about how big data changed our way of capturing, processing, analyzing, and exploiting data in new and meaningful ways for business decision makers. Now many companies undertake analytical projects at a departmental level, redefining the relationship between business and IT by the adoption of Agile and DevOps methodologies. Real-time processing, machine learning algorithms, and even artificial intelligence are the new normal in business talk.

However, companies are still struggling to adopt big data at a corporate level. In many corporations, there is a gap between launching departmental projects and industrializing and scaling-up those use cases across corporations. Embedding big data in scalable business processes is crucial to becoming a data-driven organization. Building an Analytics Center of Excellence (ACoE) can be the basis for this transformation.

Remaining challenges

There are three important issues that must be addressed in order to scale-up big data across a corporation and make a real impact on business outcomes:

1. Lack of skills across the organization.

There is an identified global shortage of analytical talent, a set of data experts ranging from data engineers, big data architects, and data scientists. It is not easy for a company to find these profiles, attract them and retain them. And it gets more difficult as technologies continuously evolve at a challenging, rapid pace. When a company employs these experts, they are not always equally distributed throughout the organization but sometimes concentrated in a particular department or business function (for example, in the risk or marketing departments), making it difficult to leverage these skills for the good of the entire organization. If a company has multiple locations, it is even harder to keep the right balance of skills in all the subsidiaries.

The shortage of skills affects the technical or analytical departments as well as the business areas. Companies need subject matter experts who understand business needs to communicate with the data experts, as well as managers able to make decisions based on data and supported by facts more than by personal, biased experience. Furthermore, the new skills require new ways of working and therefore an organizational cultural change.

2. Lack of standards, methodology, and governance

Even mature organizations with analytics teams in place in different departments, business units, or countries find that every team tends to work with their own tools, libraries, software versions, and data sets. This variety can make it difficult to industrialize and implement global solutions, and ensure code reusability. Companies need to define standards regarding coding, tools, version control, and quality control, and have all the teams working with the same tools and sharing their methodologies. Additionally, analytics teams must have big data governance policies and processes in place, controlling and limiting the access of data in the data lake and ensuring security and data privacy controls. In Europe, for instance, a new General Data Protection Regulation (GDPR) requires a very demanding process regarding data traceability at a field level. Data is a key asset, and companies will be required to protect it. big data governance is a must not only because of legislation, but in order to secure a business’ reputation for security and privacy.

3. Lack of use cases prioritization

Big data must be led by business, but often companies face organizational issues as a result of the lack of use case prioritization. Although the companies are willing to implement big data projects in various business areas, the projects are usually planned according to the big data architecture roadmap or the data provisioning strategy defined by IT, instead of prioritized according to business impact. When there are several departments pushing for a centralized big data budget, a business-oriented, ROI-driven approach is needed to define the use case roadmap in order to maximize the impact the entire organization.

Accelerating big data adoption by business

These three challenges are not an excuse for failing to adopt big data solutions. The most effective way to create mechanisms to deploy big data across the entire organization in a systematic and scalable way is to launch an ACoE.

Although the concept may be familiar, the element of ACoE we’re focusing on here is not that of an “algorithms factory” or “a technical team of product specialists.” Instead, an ACoE should consist of a team of business and technical people with centralized and distributed capabilities and resources, working in big data advanced analytics projects and creating a common workspace in which methodologies, tools, models, and techniques are shared in order to gain efficiency when implementing the initiatives across different business units and markets.

The operational model to make an ACoE successful is not obvious. Here are some of the key principles of an ACoE.

First, it must be connected with the business. The ACoE must include a team of business experts able to align the business strategy of the company through the prioritization of use cases and coordinate the implications at a technology, architectural analytics, and governance level with the different stakeholders.

Second, it has to be able to grow organically; as the company expands and scales-up new big data projects across the corporation, more people and teams will join. The AcoE hosts what we call the core team—a centralized function—while the expansion is through the extended teams, a distributed function across geographies. The core team shares and expands best practices and methodologies, accelerating know-how transfer.

Another principle is that it should be able to deploy as a service. The ACoE must be able to deploy as a service for the business units as a shared services operation, a fully outsourced operation, or in an extended capability, allowing infrastructure and resources rationalization, scalability, and elasticity. ACoE costs should be allocated as a service to the business units.

Key benefits of an ACoE

An ACoE is essential to accelerate big data adoption by business at scale. It reduces the implementation times drastically and therefore the time-to-market to deploy new data-driven products and services. It ensures best practices and methodologies are shared through different teams in the organization. An ACoE is alive, and it expands and it grows as the organization’s needs evolves, a key factor for realizing the business value of big data.

Data is the key competitive advantage and differentiating factor for companies in any industry. A true data-driven organization understands that data is at the center of any business strategy. Companies must be able to use data not only to improve decision-making and operational efficiency, but they must have the capacity to create new products and processes based on data-driven insights. In order to do so, companies must embed the necessary organizational and cultural changes at a corporate level that it takes to succeed. An Analytics Center of Excellence is an important tool for accomplishing these goals.

Article image: Ferris wheel (source: Skeeze via Pixabay).