News

Universal data literacy: The key to getting value from your data scientists

By August 13, 2020 August 17th, 2020 No Comments

Over recent years organisations have focused on creating dedicated functions of highly skilled Data Scientists and Advanced Analysts to maximise the value of their company data. It’s now the role of L&D to develop data skills across the business so employees can translate the experts’ insights into meaningful action.

 

Centralised data science teams or external consultants are often an integral part of an organisation’s data strategy, hiring in expertise to work on complex problems. However, data scientists are often removed from the day-to-day operations of the teams they are asked to help. At the same time, people within those teams may lack the statistical knowledge to input effectively into a project. 

This causes a communication gap between the expert and non-expert and, therefore, less impactful data analysis.

For example, in any given project a number of variables that are not in the raw data will affect the data analysis. This could include new product launches, staffing changes, a shift in strategy or seasonal trends. Decision makers will want to flag these variables early in the process so the data expert can factor them into their work. The analysis, accurate as it may be given the data that was available, lacks true meaning without such valuable context.   

The result is that only one in three companies (32%) report being able to realise tangible and measurable value from their data. Developing the skills of the broader workforce will close that gap and maximise the value of organisational data. 

Working with – and challenging – the data experts 

A data-confident employee can act as the conduit between their team or organisation and the data experts, communicating effectively with them and challenging them during projects. They will be able to: 

  1. Understand the high-level concepts underpinning many advanced analytical techniques
  2. Know how to input into a project brief to ensure a quality output 
  3. Be confident in challenging experts to communicate their work without statistical jargon 

Digital transformation is only as successful as the capability of the employees who are going to live it and breathe it, will allow. As part of the people development component of digital transformation strategies, key data skills need to be prevalent among employees who have access to and use business intelligence tools created by the data scientists and analystsThese skills include: 

  • Being able to have meaningful conversations about data with a range of different stakeholders 
  • Making decisions based on data rather than hunch 
  • Communicating insights clearly and concisely 
  • Understanding how humans process information 
  • Grasping basic statistics and probability 
  • Knowing about the practical implications of GDPR and data protection legislation 
  • Confidently assessing the reliability of data and other evidence  
  • Having a basic understanding of principles of machine learning and predictive models 

More than half (53%) of the working population do not have the essential digital skills needed for the workplace, and this is predicted to rise to 66% by 2030 (Industrial Strategy Council, 2019). 

Bridging the gap between data experts and the professional population will have a profound impact on the business. Senior leaders will find that data analytics projects are more valuable as more people understand and know what to do with the insights and findings. They will be able to ask pertinent questions, holding the experts to account in terms of getting to the crux of real business challenges. There will be an overall improved flow of information throughout the organisation through better communication of insights. Last, but certainly not least, there will be less chance of data breaches. 

A little knowledge can go a long way. Speaking the same language is critical for any collaboration to achieve its potential. Data literacy is no longer reserved for the elite few.