Data literacy training: What you need to know

Data literacy training: What you need to know

Successful data literacy training programs are never one-size-fits-all. Consider this expert advice to avoid common mistakes and design a data skills plan that works

up
19 readers like this
data literacy training how to design

Recently, I was talking to the Chief Digital Officer of one of the oldest banks in America, which started its digital transformation journey in early 2019. This CDO had played a pivotal role in the bank’s journey, bringing about a significant transformation in security, cloud migration, and overall master data management.

He was proud of the data team that had overhauled the bank’s legacy infrastructure. The team had transformed the systems and processes and ensured that more data was being captured and made available for analysis. Dashboards were built to offer marketing data, customer data, customer support, and operational data all in one place.

However, the bank had a big problem: End-users were not leveraging the DIY, slice-and-dice dashboards that were available. Instead, they wanted to be spoon-fed custom charts and graphs, as they had been in the past.

Fundamentally, my CDO friend’s problem stemmed from a problem at the core of the organization: lack of a data-driven culture. The organization had already achieved data maturity, but it was severely crippled by a lack of data literacy.

[ Get answers to key digital transformation questions and lessons from top CIOs: Download our digital transformation cheat sheet. ]

Data literacy training programs: 3 common problems

To achieve enterprise-wide data literacy, the bank’s head of HR was tasked to lead that part of the transformation. Seeking advice, my friend the CDO shared with me the HR exec’s plan to build data literacy for the entire organization. Her plan had a number of problems, but three stood out.

1. The same data literacy training for everyone

The plan included the same data literacy training for everyone, regardless of their role or skill set. The plan assumed that everyone – from executives, marketers, bank tellers, and claims analysts to customer support agents – needed the same level of data skills.

This is far from reality. Having led large-scale enterprise-wide data literacy for the last nine years, I’ve learned that most organizations have between six and eight distinct data literacy personas, each with its own end goals.

We have identified 24 data literacy skills and competencies. If all employees were forced into the same box, some would not reach the level they need to perform their job while others would find the training too difficult – and just about everyone would be frustrated.

If all employees were forced to take the same training, some would not reach the level they need to perform their job while others would find the training too difficult – and just about everyone would be frustrated.

Here are seven common data personas we see, in order of increasing data literacy:

Data-enthusiastic: Most frontline professionals, including bank tellers and customer support agents, fit in this segment. This persona is typically enthusiastic about data usage and understands basic charts and tables.

Data-literate: This persona primarily applies to mid-level managers. They understand the overall structured process of data science and basic statistical methods and can draw meaningful inferences from various types of tables and charts.

Data-driven executive: Nearly all executives, from the director role and up, fit this segment. Those in the category are data-literate and generally also have leadership skills that utilize data-driven thinking and hold teams accountable for fact-driven decisions.

Data-educated: This is the first level of data literacy where professionals are expected to perform hands- analysis using simple methodologies like aggregate analysis and correlation analysis. They might perform this analysis using Excel or other BI tools. Typically, all product managers and operations professionals fit this category.

Citizen analyst: This group includes professionals who are data-educated and who also have skills in hypothesis testing. Typically, marketers, business analysts, and other analysts who are not data scientists and who do not work in the data department fit this persona.

Data analyst: This group includes those with all the skills of a citizen analyst as well as data extraction skills using SQL or other comparable tools. Typically, non-data scientists in the data department fit this role. Some data analysts may be able to build simple statistical models in Python or R, for example.

Data scientist: This self-explanatory persona has all the skills as a data analyst but also has advanced data science skills in building statistical models, machine learning models, and deep learning/AI models using R or Python or other comparable languages.

2. Training is made up of MOOCs

There’s nothing wrong with MOOCs (massive open online courses). They fulfill many training needs effectively and efficiently. However, data literacy is best learned using an end-to-end approach that teaches the entire cycle from business questions to insights and actions.

MOOCs are more modular and generally don’t include key topics like what questions to ask, how to create a hypothesis-driven plan, how to engage stakeholders at critical junctures, how to prep data and systematically go through analysis to find actionable insights, and how to deploy those insights towards actions and dollars.

Remember, at the end of the day, data literacy is about learning skills that help the professionals in your organization use data for better decision-making.

At the end of the day, data literacy is about learning skills that help the professionals in your organization use data for better decision-making

3. Training does not include use cases

Another common mistake is to equate training and data literacy. Data literacy is not an easy skill to learn, and if training does not include use cases, it will likely become just a training exercise that’s forgotten by the weekend.

When I started Aryng in 2011, I believed that if we trained our team members in appropriate data skills – as we did at Google, Box, Epson, PayPal, and other companies – they would become data-literate. Although our before-and-after training survey showed high satisfaction and earned rave reviews, a few months later we discovered that much of the framework that was taught had been forgotten and was not used.

That showed us the importance of immediately applying training materials. Over the years, we also learned that mentoring employees while they applied their learnings on use cases, along with cohort-based learning, was highly effective.

To achieve the best data literacy outcome, have your employees immediately apply what they have learned to their current workflow.

5 steps to data literacy success

To ensure a successful data literacy program for your organization, consider the following five steps.

1. Map your data literacy personas

Clearly define the data literacy levels and specific outcomes needed for each persona.

2. Conduct a data literacy assessment

I recommend assessing every individual against their target persona so you know the baseline. We do this by mapping 24 skills and competencies to the desired outcome for every persona and assessing individuals on those skills and competencies.

3. Create success metrics

How will you measure the success of your program? Will you base it on project impact, learning transfer rate, or pre-post assessment delta? Consider this question carefully so you will know whether you need to make adjustments as you roll out your data literacy program.

4. Create smaller learning groups and team leaders 

With online self-paced learning paths and weekly discussion groups moderated by team leaders, you can scale learning for thousands of employees quickly.

5. Follow up training with hands-on project application 

Team leads are responsible for running cohorts and identifying and leading the projects, which should include learning, discussion, and mentored project work. Be sure to measure success metrics and modify programs as you go.

If my CDO friend is able to shift his organization’s data literacy journey to start with these key steps, his big problem will be resolved. His end users will be able to leverage their dashboards to make impactful decisions, moving the bank’s digital transformation forward.

[ Get exercises and approaches that make disparate teams stronger. Read the digital transformation ebook: Transformation Takes Practice. ]

No comments yet, Register or Log in to post a comment.
A highly-regarded industry thought-leader in data analytics, Piyanka Jain is an internationally acclaimed best-selling author and a frequent keynote speaker on using data-driven decision-making for competitive advantage at both corporate leadership summit as well as business conferences.

7 New CIO Rules of Road

CIOs: We welcome you to join the conversation

Related Topics

Submitted By Carlos M. Meléndez
November 25, 2020

Organizational buy-in is a key part of successful digital transformation. Here's how to tap the talent and perspective of some key roles as you build your strategy.

Submitted By Michael Del Castillo
November 25, 2020

At a time when CIOs can use cloud infratsructure to turn on new money-making services for customers overnight, how should should we measure IT success? Hint: It's not about uptime

Submitted By Kevin Casey
November 24, 2020

Why are the principles of the Standard Operating Environment (SOE) even more relevant in the age of containers, Kubernetes, CI/CD, and automation? Let's explore efficiency, security, and related challenges

x

Email Capture

Keep up with the latest thoughts, strategies, and insights from CIOs & IT leaders.