Philippe Nieuwbourg (@nieuwbourg) is an independent analyst, author and lecturer focusing his work on information technology to improve a data-driven economy. In today’s part 1 of this two-part interview, Nieuwbourg discusses the founding of Decideo and the evolution of data management.
Can you tell us about your professional background? How did you become interested in Data Science?
I have been working in “data” since I created Decideo, more than 25 years ago. After studying accounting, I worked in a couple of companies, first to install accounting software in medium and large-sized organizations, and in a professional IT magazine as editor in chief.
After those experiences, I founded Decideo, as the first French-speaking professional community about “data.” It has been called reporting, business intelligence, data warehouse, business analytics, big data, artificial intelligence … words change like marketing waves, but it’s always about data!
What have been the most interesting developments in the field since you started your career? What has made the biggest impact on the way we use data today?
For 60 to 70 years, Information Technology was only about numbers. Created during the second world war, the first computers for decades were just able to manipulate numbers and characters chains viewed like ASCII codes. The shift came with mobile phones, social networks and micro-processors power.
Since then, we can collect, store, manipulate and analyze what is called “unstructured data” coming from photos, audio and video files. That’s a huge step. It means that what we say, hear and view, can be “understood” by computers and analyzed, a little like our brain did it.
With the progress of artificial intelligence and deep learning, it’s clear that we will soon have an augmented intelligence available at our fingertips. Unstructured data analysis will have a huge impact on businesses during the next years.
What do businesses need to know about selecting the right tools to manage their data? How should they approach finding the tools that will work best for their needs?
Don’t focus on the data visualization part! It’s a sexy, amazing and entertaining software to choose, but it’s the cherry on the cake. Keep it for the end of your project. First focus on the data integration tools. It’s much more important. You could have the best of class data visualization tool, but if your data isn’t accurate or comprehensive, you won’t find any value in it.
During the typical day of a data scientist, only a small part of their time is used for noble tasks, like machine learning, data storytelling, graphical analysis. More than 50 percent of their time is used for manipulating datasets and fixing data quality problems. List your data sources (for today and tomorrow), think about data quality, metadata management, GDPR, regulation, privacy, security. When you will have fixed all this, enjoy a little time to choose the best graphical tool. It’s the easy part of the job.
What are the most common mistakes you observe businesses making when searching for and using different data management tools? What should they be doing differently?
The most common mistake is to focus on technology. Believe me, you can do a lot of things without having a Hadoop cluster in your data center! I’ve met a lot of companies, like a bank I remember in Canada, which bought a Hadoop distribution without knowing why – only because its main competitor did it before. And three years later… nothing… Why? Because it was a technology purchase made without connection with business needs.
What type of skills and training would you like to see more businesses focus on when it comes to data management? What training should they invest in to prepare for the future?
I think that “data analysis” skills are not an option anymore. Especially for business people. Do you really want to rely on an IT department that just focuses on technology and doesn’t really understand your business needs? I don’t. All business people should be trained to acquire basic skills on data management and data analysis. If data is the new oil of the economy, all business people should be able to generate value from it.
I don’t try to say here that IT people are not doing their job. They have the most important one: focusing on infrastructure, compliance, cybersecurity … but have to let business people take care of data analysis.
In universities and business schools, all students should be prepared to manipulate datasets.
Continue to Part 2, in which Nieuwbourg talks about what organizations should be doing with data today, and down the road, and trends in Big Data, IoT & data visualization.
Download Syncsort’s latest eBook to discover the new rules that are redesigning the relationship between business and IT.