Home » Uncategorized

Aligning Data Science and organizational structure: how companies are solving this issue?

Politecnico di Milano is investigating on it.

The proliferation of data and the huge potentialities for companies to turn data into valuable insights are increasing more and more the demand of Data Scientists.

But what skills and educational background must a Data Scientist have? What is its role within the organization? What tools and programming languages does he/she mostly use? These are some of the questions that the Observatory for Big Data Analytics of Politecnico di Milano is investigating through an international survey submitted to Data Scientists.

Nowadays the Data Scientist is a figure not always integrated in an organization, sometimes organizations needing for Data Analysis skills ask for external consultants instead of having one or more Data Scientists internally.

The organizations having internally Data Scientists can decide how to manage them choosing between four models, the “Centralized” model, the “Business Driven”model, the “Matrix” model and the “Mixed” model.

By adopting the Centralized model the Data Science teams serve the entire organization but report to a Chief Data scientist who decides which projects the teams will work on. This model often works best for organizations that are operating with limited resources and have too few Data Science experts to embed any in business units for long-term assignments.

By adopting the “Business Driven” Model Data Scientists are fully embedded in-business units such as marketing, R&D, operations and logistics.

The third one is the “Matrix” Model, the difference with the last ones is that the embedded teams in the Business Driven model report to a single Chief Data Scientist as opposed to business unit leaders. This model works best in organizations with medium-size Data Science capabilities.

The fourth is the “Mixed” model. This is more complex and it’s a useful choice for large companies, it combines the “Centralized” and the “Matrix” approaches, with a hybrid configuration. Some cross-organizational support tasks are demanded to the Data Science unit whereas other business-related responsibilities are arranged with a Matrix model.

The partial results of the survey by Politecnico di Milano show the following results.

Aligning Data Science and organizational structure: how companies are solving this issue?

Politecnico di Milano asked Data Scientists to answer about the department they’re located in. According to the pie chart it’s possible to notice that 30% of the Data Scientists interviewed work in a department located within a line of business (e.g. Marketing, Operations, Finance and R&D), that means that the company is adopting the Business Driven or the Matrix model. The 26% of Data Scientists work in an independent department, specific for the Data Science’s activities, like the Centralized model suggests; 15% is actually working within the IT department and 21% is working as an external consultant.

What emerges from the partial results of the survey is that there isn’t one model significantly more represented than others, it means that the organizations are still trying to best fit the Data Science teams in their company and that sometimes there are still not ready to have these competences internally.

Leave a Reply

Your email address will not be published. Required fields are marked *