It’s not the size of your data, but what you do with it that counts

Data is central to ensuring today’s businesses operate efficiently and effectively. But collating the data you need is only the first, easiest step – the real work begins when the data must be processed.

 

Numerous elegant technological solutions have certainly been devised to help human employees make sense of the mountains of information produced daily by any business. Data virtualisation, for example, creates a data access layer above the organisation’s disparate data sources, giving easier access to that data and enabling data silos to be dismantled.

 

Unlike traditional methods of data aggregation, this bypasses the need for additional storage capacity, or controls and governance to ensure security and compliance when analysing the data. Moreover, data can be automatically listed in what’s called an enterprise data catalogue – a repository serving as a data marketplace, where it can be searched, accessed and reused too.

 

Data catalogues also contain information that can assist users in deciding whether some data is in fact fit for the purposes of their inquiry, and have help in data preparation too.

 

As these technologies provide better insights into previously siloed corporate data, self-service data provisioning has become more feasible. But this brings with it its own stumbling block, in the form of the general business user who often feels statistics – never mind data analytics – is beyond their grasp. For things to truly take off, the human element in the data equation must equally be considered. It’s no surprise that data literacy is becoming a skill that’s increasingly in demand among any business that’s serious about making the most of its data.

 

So where do we go from here?

 

Is upskilling the right approach?

Kevin Hanegan, a founding partner of the Data Literacy Project and the author of Data Literacy In Practice, maintains that it’s top management’s lack of trust in employees’ abilities to use data analysis tools that stands in the way of data democratisation.

 

To make full use of self-service data provisioning tools, all users need to possess a basic understanding of statistics. And to keep up with today’s increasingly data-driven era, data literacy should be integrated into school curricula and listed as job requirements equal to numeracy and computer skills.

 

According to Hanegan, data literacy is about how can you maximise the transition from data to insights and values, no matter whether you analyse or consume data. However, some soft skills are also key when dealing with data, such as challenging your assumptions and mitigating your biases, as well as listening to diverse perspectives.

 

Jordan Morrow, another advocate of data democratisation, points out that “unless it is coupled with data literacy, democratising may not amount to much.”

 

Augmented data analytics

But if the human brain, as Hanegan says, is susceptible to biases when it comes to statistical data and probabilities, would it be better to entrust machines with the heavy lifting when it comes to crunching data?

 

Augmented analytics is still in its infancy, but this is exactly what it’s designed for. While the main use cases for business intelligence tools are fixed reports and dashboards, augmented analytics can enable the business user to complete more sophisticated data-related tasks, by offering capabilities such as automated data preparation, guided analysis and natural language processing.

 

The data preparation functionality makes advanced data discovery – the collection and evaluation of data from various sources – an easier exercise for lay-users, providing a step-by-step guide and recommendations to assist the easy integration and preparation of data. Augmented BI tools can also automate the extraction, transformation and loading (ETL) of data.

 

Meanwhile, the natural language generation and processing capabilities of augmented analytics platforms eliminate the need for non-IT users to have to write their queries in a programming language. These advanced analytics tools can both suggest types of visualisation most suited to the data to be presented, or – enabled by NLP – can translate findings into human speech. (According to Gartner, data narratives will start superseding dashboards as the primary format for presenting data as early as 2025.)

 

To get around in a data-driven world everyone needs to understand data

We can’t entirely discount the possibility that, someday in the distant future, anyone without any specialist background can collect and analyse data with the same ease of creating a chart in Excel today.

 

However, until a completely automated data analytics tool has been developed, we’ll still need humans with good data literacy in the loop, augmented – but certainly not replaced – by machines.

 

Considering that it’s increasingly not just workplaces that are driven by data but more and more aspects of our personal lives too, understanding data at least to a certain level will soon be a crucial basic skill – if it isn’t already.