The age of data is now, and it’s set to impact your role in ways hard to imagine. Large enterprises collect more data than ever before and keep unlocking new business opportunities. The trend is well studied, with IDC predicting revenues for big data and business analytics solutions to reach double-digit growth through 2022. While the importance of having data-centric skills and talent organised in data analytics teams is clear, other aspects of organisations tend to be omitted. Following IDC’s white paper The Digitization of the World – From Edge to Core: “Dealing with digital transformation will require not just new technology, but also new skills (…) and relationships with top management”. Each and every part of an organisation should be data-ready, as the impending change spans more than just data science teams and departments. Future success lies in integrating data-centric processes and methods across all departments to “reach new markets, better serve existing customers, streamline operations, and monetize raw and analyzed data”.
In the early 2000s, companies prioritised hard skills such as ETL development (extract, transform, load), Python, R (programming language), SQL and web development. Nowadays, many of these technologies are embedded in data platforms and organised into packages. An analysis that would require weeks just to write and debug the code now takes only a couple of lines. This natural progression resulted in a shift in skillsets desired by employers.
Employees need to create robust data governance and best practice guidelines, assure cybersecurity and privacy of data as well as build trust within the organisations – they need to exhibit good data literacy. As analytics becomes a more critical component of businesses and their future growth, professionals are faced with new problems:
- How to manage data?
- What new governance structures and processes should be put in place?
- How to manage new types of contracts and services?
- How to translate business-specific expertise into meaningful data analytics?
- How to ask the right questions?
- How to challenge the outputs of algorithms?
All of the above go hand in hand with increased data usage, bringing data literacy to the centre stage. At its core, the concept is all about utilising soft skills in data-rich environments to address business needs. From engineering or HR to environmental management and legal departments, there is no coming back from the transition.
Insufficient data literacy
The cost of not understanding the context of data is huge. A global miner’s maintenance analytics team discovered a large variation in railway quality in a very specific part of their supply chain. They studied the correlation of track vibration to various metrics and reached the conclusion that car dumpers had uneven tracks and should be repaired to avoid excessive repair and maintenance costs. When the team (which spent more than three months analyzing the data) presented the results, the executives who heard the story were impressed. They believed there was something big in the findings. Until one of the executives asked: “Do you know how a car dumper operates?” It turned out that the vibration data was sourced from accelerometers attached to train cars. As car dumpers are designed to flip ore cars upside down to unload the ore, each and every accelerometer repeatedly showed high variations in the recorded data which were inherently linked to the usual, as-designed operation of the machine.
That context had not been considered and rendered the initial iteration of the analysis fruitless. Ultimately, the findings were significant, the team went back and learned a lot about their business. The basic consulting and business skills of understanding context and creating clear definitions of data were missing. Sometimes, insufficient data literacy can not only bring teams to wrong conclusions but also prevent them from achieving any outcomes altogether.
Traditionally less numbers-driven departments such as HR, HSE or legal are a part of the data literacy bandwagon as well. Without quantifiable, measurable and understandable metrics it’s difficult to improve processes, speed up implementations and offer additional value. In the case of HR, the data-literate approach can help promote a fairer and more equitable approach for assessing employees and facilitate a more open and transparent promotion process. It allows extracting additional value by mapping skill sets to identify the most adequate profiles for hiring, or to identify complementary roles to boost productivity in teams, or areas of operation. In a recent survey performed by Visagio in Brazil, it was found that only 24% of companies have an initiative focused on Workforce Analytics. The main limiting factor proved to be data availability and quality: only 13% of companies indicate that they have access to high-quality data. As for its daily use, the survey revealed that the most common practical applications of Workforce Analytics are focused on dashboards and workforce planning, demonstrating limited applications for recruitment, performance evaluation, career succession and talent retention.
While limited growth and sub-optimal processes can have a negative impact on a business, they usually do not directly impact the bottom line. On the other hand, the cost of having misaligned data owners and data users can be significant. An analytics team in a global chemicals manufacturer aimed to study historical data to pinpoint the largest pollution sources within each plant. The insights were intended to help build a data-driven pipeline of projects and enable future capacity growth. Over the course of the initiative, the main obstacle proved to be the data quality. In the past, each plant had independently rolled out multiple air quality monitoring stations to meet regulatory requirements. Their main goal was to comply with environmental licensing: as long as the devices recorded data in the database, it was assumed to be correct. Years later, when company executives intended to roll out the aforementioned improvement initiatives aimed to leverage the large amounts of historical data, they were met with a somber reality check. The owners of the data (regulatory compliance teams) were not the users of the data (analytics teams) and did not have the capabilities to check and investigate whether the numbers recorded in their database were accurate. Ultimately, a significant portion of the data quality issues was fixed retrospectively, but the lack of basic quality checks pushed back the delivery timeline and overblew the budget.
Such cases are just the tip of the iceberg of how insufficient data literacy can negatively impact organisations. Regardless of whether a business challenge is related to maintenance, environmental regulation, legal or HR, prioritising data literacy will pay off in no time and now is the best time to start investing in it. Managers and team leaders can spearhead the change by asking just a few simple questions and implement training programs to address related knowledge gaps:
- How many team leaders understand the difference between owning data and using data?
- How many people in your business can interpret fundamental statistical calculations such as medians, averages or correlations?
- Are managers able to construct a business case based on concrete, accurate and relevant numbers?
- Are managers able to explain the output of their systems or processes?
- Can business leaders generate insights and apply the essence of the data you share with them?
The rise of the importance of data literacy is in many ways similar to what happened with the adoption of computers throughout the late 20th century. At first, they were introduced into workplaces only at certain specialised departments and as the technology developed, more teams and industries started adopting them. Now, due to the immense value they provide, it’s hard to imagine any office job without a computer. There is no way of stopping the data revolution and business leaders need to provide appropriate training and recruit employees who can fluently navigate it.
About the author
Wojciech Adamczyk is a Visagio consultant specialising in end-to-end data analytics projects: from data governance to insights generation. He is passionate about data storytelling and improving data literacy in organisations. His physics background enables him to solve problems using first principles by breaking them down to their simplest elements.