Understanding the democratisation of data

Dealing with an increased amount of data requires an adaptive, agile approach.  The organisations that succeed are those that can make sense of the data, spotting the opportunities and assessing ideas quickly.  However, before data can be used, it needs to be interpreted and understood properly. 

Encouragingly, with technology progressing just as quickly as our data universe has expanded, making better sense of data – at scale, and at speed - is no longer out of reach.  And the rise in advanced analytics has played a huge part in supporting this.  Analytics gives you the power to connect data, to understand it in huge volumes, across systems, and deploy it in decision making through advanced automated technology.  Most importantly, advanced analytics is today not restrained to tech companies, but available to every business in many ways.

Our research suggests that some are grasping these opportunities better than others, with 35% of businesses being able to use analytics to extract useful insight from their data-set.  Understandably, others are feeling quite overwhelmed.  Four in ten businesses admitted they were struggling to cope with the volume and complexity of data.  Half said they couldn’t use their data to drive decision making, while 71% named enhancing their analytics capability as a top priority for their business.

At Experian DataLabs we understand the importance of extracting the hidden value from data.  Our data labs offer a risk-free environment devoted to identifying new ways to glean greater insights, developing innovative solutions to today’s most difficult challenges faced by businesses.

Big data, big challenge

So, what’s stopping organisations from unlocking the insights held within their data?

Many businesses are simply overwhelmed by the sheer volume of data at their disposal.  Data quality is an issue too, with most analytical teams spending too much time on basic data cleansing and management as opposed to more sophisticated tasks that build on this.  Disparate systems and processes don’t help, which means analytics are only being applied against specific tasks, as opposed to holistically.

Once you solve the problem of poor data quality, it’s important that this is maintained.  Data decays and new insights become available, so it’s vital that organisations have the functions in place to preserve an accurate and up to data view of data in a way that fits into business-as-usual processes.  Through more holistic and reliable customer insight, organisations can make more intelligent, data-driven decisions and stop relying on subjectivity to make decisions.

And a lack of investment into the right places may also be holding up many businesses getting a better understanding of the data they hold.  There is an opportunity with the right analytics technology to create and orchestrate a data insight strategy that benefits everyone and underpins an entire organisation, giving broad business advantages.

For many businesses, the challenges outlined won’t be anything new and most will be, or planning to, take appropriate steps to move closer to capitalising on the huge opportunities presented by the growth in data.  But as the focus continues in these areas, are businesses becoming too over-reliant on a small number of employees who specialise in data?  For non-data scientists, even the basics of data can be far beyond their comfort zone.

Achieving data democratisation

Following a heightened focus on data, the employment marketplace responded by combining aspects of the long-existing professions of actuaries, software developers, business-intelligence analysts, and consultants into the newly emergent role of “data scientist”.  In the few companies that data-science practice has been happening, it has been done in a low-level, very labour intense fashion, with expert data scientists being intimately familiar with their modelling algorithms and their quirks.

In the last few years, certain parts of data science have been automated and systematised.  The introduction of several new platforms has promised to fully automate the next generation of machine learning models.  Although the first generations of these platforms would require considerable human contribution and expertise, newer models have moved towards more automated modelling, assessment, and, in some cases, deployment.

Today, data gathering and storing is a lot easier and cheaper.  Modelling algorithms are adequately powerful and push-button software applications are promising to solve all modelling needs.  With the increased use of automation techniques to understand data, it’s easy to assume that human intervention will, at some point in the not too distance future, no longer be required.  This is far from the truth.

While the skills gap still a well-reported problem, and businesses struggling to build teams that can spot the opportunities in data, it’s easy to rely on the advances in technology to address this on-going recruitment issue.  This, however, would be the wrong approach to take, as it’s hugely unlikely that any skills shortfall will be filled by machines, rather new jobs being created to support data architecture and data insight strategies.  The need for data expertise in your organisation will, if anything, become greater.

The best way to enable fast discovery and deeper insights is to disperse data science expertise across the organisation.  That, combined with easy access to analytics, can empower non-data scientists with a more purposeful tool kit for the task at hand.  In most businesses, data scientists are required to complete the most basic of data related task.  By creating a culture where sharing data knowledge and tools is the norm, these responsibilities can be distributed across a data-literate workforce, freeing up valuable time for your data specialists to be truly innovative.