3 steps for creating a data-to-value ecosystem

Although many organizations are using artificial intelligence (AI) and machine language (ML) tools as core enablers in their data analytics projects, and AI spending worldwide continues to rise, the hard truth is that most data science projects are doomed to fail.

There are several reasons for these failures, ranging from the inherent complexity of AI/ML initiatives and the persistent lack of skilled talent to challenges that exist in data security, governance, and data integration. These issues are collectively referred to as concerns for” data readiness,” according to an IDC global survey of more than 2,000 IT and line-of-business decision-makers, all of whom are involved in some level of AI use or development.

Making matters worse, while most companies routinely maintain large amounts of data, it is often stockpiled in functional silos and not easily accessed or used across these boundaries. Advances in cloud computing, data engineering tools, and machine learning algorithms are also coming faster than products and new processes can be deployed. Then there are the competitive challenges that come from both traditional channels and new, disruptive technologies.

To overcome this reality and create new value for customers and shareholders, IT leaders must create a community and culture that can accelerate and sustain the growth of data science and analytics throughout a company.

A critical need for competency

It is obvious that data science requires skills in computer programming, data engineering, math, and statistics. What differentiates a good data scientist from a great one, however, is the ability to translate the business requirements of functional domains across the company. These resources can be scarce, so they must be focused on the projects with the highest ROI/quickest time-to-value ratio, while at the same time helping to grow the overall analytics community within the company.

Quite often, a data science center of competency (CoC) is established as a resource for the company to achieve these goals. Traditionally, these CoCs reported up through the IT organization because of their codependency on the data infrastructure.

However, while data architecture and governance are critical, a data science CoC that is too far removed from the business domains will result in misalignment of goals, delays, and ultimately, increased project failures.

A hybrid organization approach, where a focused data science and analytics center acts as a bridge between the traditional IT infrastructure and the functional domains, is essential to enable success. This can be successful in accelerating the development of the data-to-value ecosystem and the culture shift necessary for sustainable growth.

Setting sights on ROI success

The following foundational goals are critical for developing a strategic plan and process for competency and sustainable enablement:

1. Create a center of competency. Projects often fail because they are developed in isolation, without consideration for the entire lifecycle of the model as well as digital thread, lineage, and data pipeline requirements. People may hold onto or hide data and information out of a belief that it may help them personally. This attitude impedes the potential for creating value when looking for deeper insights.

It’s important to understand that data science and analytics are a team sport. Creating a center of competency that focuses on collaboration, education, and inclusion will help build trust between functional organizations.

2. Extend data and design literacy efforts. Create a virtual community across the entire enterprise for fielding questions from the most basic concepts to the most complex data science and design thinking constructs. As part of a CoC, this resource hub will drive development and administration of curated plans of study, ranging from the “onboarding analytics” skill level to more advanced data science certifications.

This hub will also be a focal point for the training and certification of new data scientists within all the functional domains across the company. The goal is to create a cross-functional community that provides support for everyone in their data literacy journey.

3. Create a cross-functional and diverse team of strategic thinkers. This provides a company-wide platform for sharing ideas and identifying projects with the highest potential. It also enables team members to leverage each other’s skills and domain knowledge to co-create new value for customers and shareholders.

The alignment between strategic KPI’s and individual metrics reduces the friction for a desired culture shift and puts a focus on the highest ROI projects. Ultimately, however, a data-to-value ecosystem is not sustainable unless trust is established—in both the integrity and security of the data pipeline and between people and across functions. Once trust and alignment are established within this broader community and project funnel is established, the goal of driving value into the business rate can be achieved.