The Data Skills of Tomorrow

Key Insights
Here we analyse the specific skills and capabilities for the Data population, their strengths, weaknesses and the Skills of Tomorrow.
Download the Report
The Data Population as a whole
Much like the marketing population we found that questions on their ambitions for capabilities consistently scored highly, systematically scoring over 3 out of 4.   
 
Also scoring highly were questions regarding data visualisation and working in a data-led way .   

Similarly to the marketing population, when it comes to the questions on the reality of their skills - their performance over their ambition, respondents systematically scored below 3.

Ambition

Questions on ambition score consistently high, with an average of above 3.

Visualisation

Data Visualisation and being Data Led score highly (2.73)

Reality

Questions on performance score systematically below 3 
Common Weaknesses
There were two key areas of weakness amongst the data population.
Area of weakness #1.
Each person in my team shares at least 1 new learning per week
Average
1.9 / 4
Area of weakness #2.
Setting up Machine Learning pilot projects
Average
1.92 / 4
Data Talkers vs Doers
We found the same correlation in the data population between the ‘Doers’ and the ‘Talkers’.  
 
Again, there was a clear distinction between those with high ambition
and a high reality in their capabilities, and those who have high ambition
but lower reality of capabilities.

The Talkers
High ambitions
Low reality of skills
The Doers
High ambitions
High reality of skills

Talkers vs. Doers - The Key Differences

We analysed the most acute gaps in capabilities between the Doers and the Talkers (score out of 4).   

We found the following showed the greatest variance between the two populations: 
Analytics and ML capabilities
Doers
2.91
Talker
1.41
Having a clear data strategy
Doers
3.12
Talker
1.62
Creating simulations that provide the business with different likely scenarios
Doers
2.70
Talker
1.23

What the Marketers say

Finally, we asked the marketers for their input on capabilities.
We asked 242 data professionals
Are there any capabilities which are important to you or your team that you would like to add?
Last Mile problem
Connecting projects to the structure and processes of their companies.
Finding the right stories to connect with their team
We asked 242 data professionals
What are the biggest challenges you're facing in the coming year?
Covid-19
Standardising processes
Using storytelling and visualisation to convince their team
The Data Skills of Tomorrow
Move from a talker to a doer by addressing these key skills of tomorrow:

01. Adopt a Growth Mindset

Much like the Marketing population, we found the same distinct correlation between the Talkers and the Doers. Those who have high ambitions, and those who have high ambitions and actually achieve them. 

In order to continuously learn and adapt and achieve your ambition, adopting a Growth Mindset is a fundamental step. Individuals who believe their talents can be developed through hard work, good strategies, and input from others have a growth mindset. 

They tend to achieve more than those with a more fixed mindset - those who believe their talents are innate. This is because they worry less about looking smart, or talking up their abilities, and they put more energy into learning. 

When entire companies embrace a growth mindset, their employees report feeling far more empowered and committed; they also receive far greater organisational support for collaboration and innovation.
Why waste time proving over and over how great you are, when you could be getting better? Why hide deficiencies instead of overcoming them?
Dr Carol Dweck - Author, Mindset

02. Improve Machine Learning & Analytics Capabilities

Machine learning and analytics skills came out as key skills that Data professionals need to excel in (pun intended).   

Machine Learning is now so pervasive that you probably encounter it dozens of times a day without ever realising it. Think of Netflix, Amazon, Youtube, Spotify, they are all using Machine Learning to power their recommendation engines.    

In data science, an algorithm is a sequence of processing steps. In machine learning, algorithms are 'trained' to find patterns and features in massive amounts of data in order to make decisions and predictions based on new data. The better the algorithm, the more accurate the decisions and predictions will become as it processes more data.  

There are four basic steps for building a machine learning application (or model).
1
Select and prepare a training data set
2
Choose an algorithm to run on the training data
3
Train the algorithm to create the model
4
Use and improve the model
Where to start?

We recommend starting with small pilot projects and your business objectives to ensure your machine learning efforts are relevant and can demonstrate value to your business. You don’t need complex, big data to start running your own models, and there are plenty of tools out there than you can use without coding.

From here you can build a data strategy that incorporates machine learning to drive outcomes.
What's the difference between A.I., machine learning & deep learning?
We discuss Machine Learning in more detail here!
How do you prepare data for machine learning & A.I.?
We discuss how to prepare data for Machine Learning and AI, so you can begin to get started today!

03. Share learnings. And learn through Simulations and Machine Learning Pilot Projects

Following similar results from the marketing population, we found that data teams need to spend more time sharing learnings between team members. The value of peer-to-peer learning is huge and represents a huge economic loss for organisations that do not foster this way of working.  

Also important for data professionals is to use machine learning and simulations to provide the business with insights to inform decision making. Making sense of what the data is telling you is the key skill, and sharing these learnings with your teams and relevant stakeholders will be of huge long term value.

Utilising Machine Learning and Simulations will also bring huge value. Accenture (2017) estimates by modeling that AI could double the annual GDP growth rates by 2035 and increase productivity by up to 40%.

04. Create a clear Data Strategy

As a result of greater awareness around data and AI, many established companies have started data and AI projects with great expectations for huge transformations and to attract star talent.  

But as we found, there is still an issue regarding having a clear data strategy within data teams. We explore why, and what you can do to solve this.  According to a study, 70% of companies globally are currently working on getting the first AI deployment operational. Yet despite all the pilot projects, large scale business transformation has not taken place.

The reality is that there are no shortcuts. Amazon, Google, Apple, and Facebook all used very different business strategies to gain their current market dominance, but their common success is down to  understanding the value of data and positioning themselves early. They placed continuous emphasis on human capability building, alongside developing, testing, and deploying the top technologies internally, so that they could offer the best to their customers.What is required a clear buy in from leadership, and a clear Data Strategy. We suggest the following steps for your Data and AI strategy:

Building a winning data strategy

The reality is that there are no shortcuts. Amazon, Google, Apple, and Facebook all used very different business strategies to gain their current market dominance, but their common success is down to  understanding the value of data and positioning themselves early. 

They placed continuous emphasis on human capability building, alongside developing, testing, and deploying the top technologies internally, so that they could offer the best to their customers.

What is required a clear buy in from leadership, and a clear Data Strategy. We suggest the following steps for your Data and AI strategy:
Translate your business and digital strategy into your data and AI vision, highlighting the biggest opportunity areas for optimising your current business. As well as looking at new innovative businesses utilising AI & data.
Identify the business processes (product development, production, sales & marketing, supply chain, pricing, HR, finance, etc.) where you want to use data and AI.
Understand the current state of your data and AI capabilities.
Describe the desired result for your business processes once data and AI capabilities have been deployed.
Define new data-driven business and product ideas.
Define your execution roadmap, including investments.
Execute the first data and AI use cases by creating your AI playbook, aiming at production readiness.
Automate and scale up operations.
In the future we'll be so good at visualising and showing the numbers, that at some point, it will be automated. It will be available directly for the stakeholders on a dashboard and all the data will be already described, automatically. That means that we would be able to dedicate more time to extract the meaning, extract the insights, and start being a bit more prescriptive, as opposed to descriptive. That's usual in an analytics roadmap.
Director, Global Digial Analytics at Adidas

Want to know your team's capability baselines?

This lite version of our capability analysis software is designed to help you understand your teams growth capability gaps. It takes around 15-20mins.
Take the Scan!