The importance of data literacy in analytics success
Jan de Villiers, Head of Cloud Academy at PBT Group
Data and the concepts surrounding it usually don’t fit nicely into the little boxes we tend to create for it.
To see this in action, just take a look at the term data literacy and search for the definition of it. You will get a variety of definitions. Here are a few that resonate most with me:
“Data literacy is the ability to read, write and communicate data in context, including an understanding of data sources and constructs, analytical methods and techniques applied, and the ability to describe the use-case application and resulting value.” – Gartner
“The ability to read, write, analyse, communicate, and reason with data.” – DataCamp
“The combination of skills and mindsets that allows individuals to find insights and meaning within their data to enable effective, data-informed decision-making.” – The Data Literacy Project
This effect of being difficult to pin down is amplified by the explosion of data-related technologies, techniques, methodologies, and languages to deal with the growing volume, variety and velocity of data. Even this list of “Vs” related to data has grown over the past few years. In that spirit, I might therefore be colouring outside of the lines you see as data literacy as I tackle the topic and its importance, in this piece.
How data literacy helps develop a data-informed organisation
A data literate workforce empowers an organisation to become more data-informed and realise the value from data analytics investments.
A recent study found that business leaders and employees predicted that data literacy will become the most in-demand skill by 2030. 89% of executives already expect all team members to explain how data informed their decisions. Investing in data literacy should contribute to improved staff retention, as 35% of employees surveyed reported they had changed jobs in the last year due to employers offering insufficient upskilling and training opportunities. According to a McKinsey survey, data-driven organisations are 23 times more likely to acquire customers, 9 times as likely to retain customers, 19 times more likely to be profitable, and 2.6 times more likely to have a significantly higher return on investment. It is therefore vital for organisations to become more data-informed, and by implication, their employees more data literate.
One of my favourite mental models to communicate the importance of data literacy and being data-informed is from a presentation on the OODA loop, a concept developed by the US Air Force. It describes the iterative cycle fighter pilots use to make decisions in combat: Observe-Orient-Decide-Act. The model applies to the competitive business environment too. The key is for organisations to have an OODA loop that is consistently smaller than that of their competition. For a company to become truly data-informed, it must embark on a journey that works towards a goal to instil a greater appreciation and passion for data across the entire business. This may require it to change its organisational culture.
Managing challenges
To realise the benefits of being data-informed, there are some very real challenges to overcome. I’d like to focus on one of the challenges that is perhaps not often considered – the nature and basic characteristics of data. A very experienced data modeller I worked with, Joe de Beer, once did a presentation where he showed a drawing of a yacht, with the title: ‘This is not a yacht’. Next, he showed a photo of a yacht, again with the title: ‘This is not a yacht’. While we were puzzled, the next slide brought everything into focus. Those were models of a yacht, mere representations of a real-world thing. His point was about how data models attempted to capture important and relevant aspects of data – the essence. However, this could easily be poor representations, and would always be incomplete.
I realised that data is very frequently just a representation of something else. For example, it could be a record of an event, but it’s not the event itself, a record of customer attributes (at a point in time), but it’s not the customer. One of the biggest challenges when dealing with data is that it is never a truly complete reflection.
To make it worse, data can be duplicated and modified incredibly easily, yielding many representations, each varying in its accuracy in reflecting the original. Historically, this duplication was a necessary part of working with data due to technological constraints. However, it’s even worse today with the proliferation of technologies, techniques, and demands on data.
Often what is considered as the original is still just another representation, and it too could be a very poor representation. Garbage in, garbage out is still very much relevant. Data literacy empowers people with the necessary awareness and skills to mitigate and ideally progressively remedy the situation.
Who champions data literacy?
While executive sponsorship is needed to ensure long term success, champions should be distributed as widely and ‘evenly’ throughout the business as possible. It’s important to identify these champions, connect them, give them a mandate to take their colleagues with them on the journey, and provide them with the necessary support.
Whilst the term data literacy is still fairly young, it is an umbrella term that includes many skills that are part of more mature disciplines, e.g., Business Intelligence, that have been around much longer. Some practitioners in these disciplines may not even think of themselves as being data literate, but they have a wealth of relevant experience to tap into. A key difference however is that while these disciplines involved specialist roles, data literacy involves a shift to empowering a broader audience with these skills.
What a data literacy transformation journey should look like
A data literacy transformation journey is about changing organisational culture. While technology has an important role to play, it shouldn’t be the focus. Technology is a multiplier, of good or bad data culture and practices. A transformation journey should be realistic, understand the organisational culture, and how to change it. The scale of the effort should not be underestimated.
Data literacy should be embedded as part of an organisation’s DNA to the extent that it can continually be replicated through the seasons of the business, as people come and go, as processes and systems change. Use whatever vehicles work for your organisation to achieve this cultural change.
While the scale of the transformation journey is significant, it should be approached in an agile manner, starting small, learning lessons, and then scaling as required. It is also advisable to use reference frameworks as an outline or map to ensure proper coverage (over time) of important elements, to guide efforts and provide categories to measure maturity and progress. Some frameworks I find valuable in this regard include the Periodic Table of Data Literacy Program Elements and the Change Management Framework.
Joint responsibilities
The responsibility to become data literate doesn’t reside solely with the organisation. Every individual has a role to play. They shouldn’t wait for formalised programmes or top-down initiatives. Each employee can start with upskilling themselves. When working with data and metadata they should consider the implications of their interactions with it, ask critical questions of the data and the processes behind it, identify assumptions and bias (including their own), consider original intent and identify other possible interpretations. Lastly, they can encourage colleagues to also work towards more data-informed decision making.
