Our organization is planning on increasing our budget on AI/ML tools this coming year, what do we need to know?

“AI is not enough. The AI/ML revolution is here. However, very few IT professionals feel comfortable using the tools. This lack of knowledge presents an apparent disconnect and supports the need for further education, or outsourced support from skilled experts. It is not sufficient for companies to solely utilize the newest techniques, nor is it sufficient to blindly use these techniques as it is not sufficient for upper management to be cheerleaders. However, that is not the whole story, nor is it the most crucial idea. How the organization as a whole uses its data is of the most importance.”

How does an organization fully utilize its data?

“An organization needs data leadership — not just leaders who are evangelical about data for business practices. Leadership that provides a clear path and enables their workers to perform their best. Managers rely on data insights to make informed decisions. Across an organization, data and analytics are deeply embedded in our processes and decisions. Business units must have a centralized data repository where they have a sense of data ownership and proactively manage data quality. Access to data must be ensured for those who need it. Companies need a Chief Data Officer.

The CDO must champion data literacy by upgrading and maintaining core analytics talent within an organization. They need to build a self-service analytics culture. All decision-makers, including senior management, should be data literate and have a firm grounding in statistical inference, experimental design, and data analytics practices. Analysts spend the majority of their time on ad-hoc analysis, data discovery, and forward-looking analytics such as predictive modeling and optimization. Business Intelligence tools will serve standard data discovery work with a SQL interface supporting all other ad hoc queries.

The CDO office should frame the problem and look at the big picture by creating clear objectives defined in business terms within the data-driven organization. The key result has to be measurable. An example of these clear and measurable objectives is: We are going to increase data literacy this quarter; we will accomplish this by implementing an Analytics Professional Development Program for our employees, we consider this program successful if we have X% of our employees complete the program. These objectives should be easy enough that in the end, you can look, and without any arguments: Did we do that or did we not do it? Yes? No? Simple.

There needs to be a clear, commonly understood vision of where the organization’s analytics culture is heading; this vision must be shared across the organization to provide teams with visibility of goals to align and focus effort. Data-driven organizations must have an objective—a clearly defined goal—and one or more key results—specific measures used to track the achievement of that goal. The purpose of the CDO is to determine how to achieve objectives through concrete, specific and measurable actions.”

So why should an organization’s employees buy into the CDO’s data-driven vision?

“Data-driven organizations have higher output and productivity than their counterparts. They have higher asset utilization, return on equity, and market value. Attrition; centralized and supported analysts are less likely to leave. Furthermore, measurable objectives delivered by the CDO increase transparency in the organization and allows employees to see a valuable and viable career path in the age of AI/ML. If teams have a better grasp of data analytics and have at least one member who is skilled at SQL, they will be more self-sufficient. That translates to more independent, agile, and scalable workflows. Finally, by making decisions that are both quantitative and qualitative, we take the guesswork out of customer reactions. Organizations can scale and innovate rapidly.”