Only 38% of employees say they share insights outside of their own department.1 What’s the reason for this? Which levers can (and should) be pulled to democratise access to analytics data within your company, across all business departments? Cédric Ferreira, AT Internet’s Director of Product Marketing, explains his vision of the “data democratisation” concept, and shares advice on implementing a secure, efficient data democratisation strategy within your organisation.
How should the concept of digital analytics democratisation be defined?
Outside of its political meaning, “democratisation” is the process that permits an activity to spread to the largest number of beneficiaries. While we frequently talk about the democratisation of certain sports or means of transportation, the democratisation of data is also a frequent topic in the business world. The democratisation of digital analytics is one of the tangible cases that today’s companies must tackle – especially when those companies are undergoing a digital transformation.
To draw a parallel with the IT world, early usages of computer science were mostly reserved for scientific calculations, and the discipline later worked its way into business administration. While computer science started out unknown to most, its field of application expanded over time, and it progressively penetrated all industries. Since the 90s, the democratisation of computer science has profoundly shaped our society on political, economic and social fronts.
Which factors are contributing to this trend?
Much like with computer science and IT, for a very long time, analytics tools were reserved exclusively for technical people. But the importance of digital in a company’s strategy progressively extended the reach of these tools, and more and more business users now want to access data (product teams, finance, marketing, sales, management…). At the heart of it all are analysts, who today must field rapidly growing volumes of requests, which ultimately encroach upon their main mission: data analysis. Faced with this bottleneck, business users feel increasingly restricted in their ability to quickly get the data they need, in an adapted format. In a context as tricky as this one, democratising digital analytics is a major opportunity for business users to gain more independence and quickly access the data they need.
What should you be careful of when implementing a democratisation project?
First, it’s necessary to supervise and guide this democratisation within a company. The handling of (sometimes highly sensitive) data must absolutely be subject to a data access and management policy which addresses all aspects: security, privacy, responsibility, processes, and rules of use. Additionally, the complexity of the digital analytics discipline and the diversity of employee profiles within a company can make democratisation a tricky task. But new developments in analytics tools, a growing level of business user maturity, and the rapid arrival of artificial intelligence are all helping to accelerate democratisation.
How can artificial intelligence help this democratisation?
AI will transform the analytics practices we know today, in terms of both data exploration and decision-making. Data science and machine learning are enabling us to unearth a goldmine of information which often remains buried in massive volumes of data – this includes anomalies, user groups displaying similar behaviours, and, of course, predictions on how key measurements will evolve. But we must beware of the “black box” effect which can limit our understanding of recommendations, and which can also result in data users being less involved in data analysis. If a machine takes the analyst’s place, this can create a number of problems, and even more in analytics where context is key. In short, humans must remain at the heart of a data democratisation approach, otherwise it’s no longer “democratisation” but rather a tyranny of algorithms.
Learn more in our free guide, Democratising Digital Analytics, and get our recommendations for establishing and implementing an efficient analytics data democratisation strategy.
1source: Dun & Bradstreet 2016