A Study by Volantsys Analytics & George Brown College, Part 1: Introduction

In a world where companies are continually engaged in a rapid technological arms race, it is no secret that mastery over data science plays a critical role in improving organizational effectiveness, reducing costs, optimizing processes, and enabling informed business decisions. Although data science promises to deliver such outcomes, many data science projects not only end up failing, but do so only after expending significant amounts of time and resources. 

Recently, Volantsys Analytics has partnered with a team from George Brown College on a study to better understand the main pain points, common challenges, and successful characteristics of data science projects. In doing so, we hope to elucidate how data science teams might be able to improve their project methodologies to achieve better outcomes. Our findings will be shared over the course of six blog posts, with this one being the first, so follow us and stay tuned for all our information!

Research Methodologies

For our study, we leveraged two methodologies which included an online survey hosted on the Survey Anyplace platform (https://surveyanyplace.com) as well as interviews with data science subject-matter experts and professionals.

The survey consisted of four parts and 23 questions. The first part focused on general information about the respondents to provide an understanding of industries and applicable approaches to data science projects. The second part collected data on approaches and tools, while the third part looked at roles in a successful data science project. The final part of the survey addressed the effective communication and collaboration within the data science implementation teams. 

Respondent Information

In the first blog of this series, we will provide a quick glimpse into the features of our survey respondents. On the whole, our 17 respondents represented a community of data science professionals or experts who rely closely on data in their everyday activities.  

Overall exposure and experience with data science was high. 94% of respondents had been directly involved in data science projects, with 30% of respondents having 2 or more years of experience in data science development. About 60% of respondents participated in at least 5 successful data science deployments with 29% of respondents involved in 30 or more successful data science projects. 

Most of the respondents represented companies working in information technology (35%) or finance and investments (29%) industries, and more than 75% of participants worked in midsize businesses (101-1000 employees) and large businesses (1000+ employees) that operate on a national or global level. Our respondents therefore provide good insight into some of the industries that tend to be more heavily reliant on data science. 

Lastly, our respondents generally agreed that the main purpose of data science projects is to introduce better evidence-based decision-making and business processes (71% of respondents). Likely secondary benefits include optimizing costs (65%), increasing revenue (59%), and customer retention (53%). Interestingly, the results imply that data science has greater applications in cost reduction rather than revenue generation, although data science likely serves a fundamental role in both. 

Be sure to join us in Part 2 of our series, as we analyze the survey results on data science methodologies and approaches!