The Importance of Data Democratisation – Q&A with Pavel Dolezal

data democratisation

 

6 MINUTE READ

 

Much of today’s attention is focused on AI and how it will affect the future of business. Even in the data scene, people are easily distracted by discussions about the latest deep learning algorithms and specialised apps. This means much less time is dedicated to the topic of data management and how to use the insights gleaned from data.

 

Pavel Dolezal, CEO of Keboola, goes against the grain. He doesn’t care so much about the latest news of an AI beating the world’s best Go player, he’s more passionate about how data can and has changed how companies do business. Keboola is a Data Value Platform, the first system that enables technical and business people to collaborate effectively in the same environment.

 

Pavel has been building companies since 1999, including Ataxo, one of the EU’s largest digital agencies in the performing media space. In addition, Pavel actively invests in early-stage tech companies in the EU, Brazil and the USA. He now leads a US office based in Chicago.

 

We sat down with Pavel recently to have a chat with us about some of his passions – data and machine learning.

 

Why is it essential for those who work in data to focus on building their organisational knowledge rather than strengthening their technical ability in machine learning?

 

We often read about new mind-blowing machine learning applications, but these are the icing on the cake. For 99% of companies, there is simply greater business value in optimising the organisation as a whole rather than in any narrow application.

 

Most companies make a profit by having people working together; it is a complicated and complex mechanism where each part has its own role and value. Deep learning and “AI apps” are only making one cogwheel better.

 

On the other hand, these two are the same problem. To not only test but to use things like machine learning, companies need to first manage their data properly. And all the data and insights machine learning can provide are meaningless without further action, which is why it is so essential to teach organisations how to fish before they buy a new hyper-modern fishing rod.

 

At Keboola, we have identified four steps on the data maturity journey that most organisations go through. The first three steps in this journey are oriented towards aligning the organisation and teaching people within it how to properly use the results from data-oriented activities. Only when the organisation as a whole undergoes this transformation does it make sense to apply machine learning on the company-wide level.

 

What are the main reasons that current methods of organisational design are falling behind the needs of modern-day businesses?

 

Without a doubt, one of the most significant obstacles of innovation we see in current organisations is information silos.

 

Almost every department in all successful companies is collecting and reporting some sort of data, at least in the form of basic KPIs. Reporting and KPIs are nothing new; they are proven and centuries-old concepts. What has changed is the exponential growth in the data sources businesses can and want to work with.

 

This technologically-driven trend bred effectivity, which in turn caused the KPIs of every single department to rise as well as the number of new data apps. But relative to that, there has been very little progress on bringing the company data together.

 

One of the biggest reasons for this is the historical debt that most companies carry on in the form of legacy infrastructure which was setup that simply doesn’t support the modern way of collaboration. That’s why we have built Keboola to be as universal as possible, platform to connect even the most prehistoric legacy systems with the cutting edge data apps.

 

The obstacle of data asymmetry has become a considerable problem when addressing mission-critical initiatives (which are prioritised below company goals, but above department goals). By definition, they need two or more departments working together to tackle them, which usually means that silos located within the departments become a problem, and most of the data apps fail no matter how much “AI” is inside them.

 

How important is the “democratisation” of data beyond those with technical skills?

 

Very. For most business applications, you need someone who is skilled, smart, specialised, and experienced to look at the data, make sense of it, and decide what action to take. To be most effective, you want to give everyone an opportunity to work with data. Since different organisation members are experts in their own field, they are able to give context and are paid to act. To empower them, we just have to provide them with better tools.

 

Data can only give you a starting point. Instead, all significant decisions in most businesses are rightly still being carried on by people. If they don’t have access to data or knowledge about analytics, how can they process the data to make the right decision?

 

How can data professionals aid in the process of “democratisation”?

 

Welcome it, take it as theirs, and lead the way. I think we have seen a similar development in IT. Historically, you called IT when you needed your computer fixed, they were considered as utilities. Later, IT became an integral part of the company‘s business, and now with digitalisation we see CTOs on the similar level of importance as CEOs, then becoming CEOs, etc. I think the same will happen with data as CIOs are beginning to recognise the real power of data.

 

Currently, many companies still view data and business intelligence departments as teams that take care of infrastructure, fix things, and solving pretty boring business inquiries. In more progressive companies or marketing departments, you will find many curious people wanting business intelligence department to give them access to all kinds of insights. Having a self-service infrastructure helps data professionals the most because it unties their hands and lets them work on fascinating projects and more complex business questions that nobody else in the organisation can tackle.

 

To further the goal of data democratisation, data professionals should choose their tools and strategies based on what will help bring together different departments across their organisation. They should also make sure their staff understands the real business value of data so that their people start really utilising it and even become enthusiastic about it.

 

What are some of the key lessons data professionals should keep in mind if they want to bring about real change?

 

Even though I love data and fight for a data-centric approach, I do not believe that data is everything. It is not a holy grail. There are other super useful and practical business tools that people need to be using. Again, for most business applications, you need people who are skilled, smart, specialised, and experienced to look at the data, make sense of it, and decide what action to take. For this, you’re going to need to get these essential people within your organisation to see the value that this data-driven approach will bring them so that they will willingly get on board.

 

So my number one advice for data professionals who are striving to bring about real change would be this – start with yourselves, learn to communicate the importance and business value of data and your work.

 


 

Planning to expand your tech team? Quick submit your vacancy to start your search today.