“It is a capital mistake to theorize before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts.” – Sir Arthur Conan Doyle, Sherlock Holmes
It is amazing how even a century later the words from literature’s most famous detective continue to ring true. Data, after all, is becoming the lifeblood of organizations globally. Organizations are moving towards becoming more data-driven to optimize their assets and improve their growth opportunities.
Given the massive volumes of data being generated, we are all willing to be led by data. But is generating the data enough? We all know the answer to that. NO.
You might be busy generating data, but you also need to create roles within your organization to push you in the direction to becoming data-driven.
Gartner estimates that “80% of organizations will initiate deliberate competency development in the field of data literacy, acknowledging their extreme deficiency.”
And while there is enough talk about the technology aspect of data, we need to focus heavily on the ‘people’ who will be playing with this data. In other words, what are the different roles (Let’s put aside the role of the Chief Data Officer. That’s a given) that power a data-driven organization? Who are these people who will ‘speak’ to your data?
What Hadoop is to Big Data, the Data Scientist is to a data-driven company. But while having a data scientist is essential to be data-driven, just hiring a data scientist does not make you a data-driven company. It’s just having a driver’s license doesn’t make you a Formula 1 racer. Or being a Formula 1 racer guarantee that you’ll be a good driver on open roads, right?
So, what kind of data scientist do you need?
Your data scientist not only has to capably apply advanced statistical models to convert data into information but also should have the capability to make magic happen with data. They have to explore and experiment for improvising Machine Learning algorithms. They should be able to production-build models easily and be willing to creatively experiment with machine learning. These are the rock stars who will grow your business by transforming, processing, modeling structured and unstructured data.
Had, Charles Joseph Minard, the famous French engineer been alive today, I think he would have made a fantastic data artist. Just take a look at the visualization that he created for Napoleon’s 1812 Russian campaign. While the expedition ended in disaster, the map is considered one of the best statistical drawings created.
The data artist, as the name suggests, is the person who presents the information generated by the data scientist in a manner that the user can understand it. Think beautiful and meaningful visualizations that make complicated information seems simple and implementable. These are the superheroes who bridge the gap between IT and business creating visualizations from obscure data that can be understood easily by business users and facilitate decision making.
Data artists are part programmers, part visual artists, and part visualizers. They are the one who helps us analyze ‘what if’ scenarios that we face on a regular basis. As they need to work closely with data scientists, data literacy is key for these superstars. And along with this they also need to be clued into human educational psychology to understand visual processing capabilities.
If you want to be a data-driven organization, you have to cultivate your own set of data enthusiasts. If you look closely, you will find this breed of people existing in your organization. Who are these data enthusiasts? This set of people, much like Holmes, want to use data to come up with business strategies, design growth plans, or offer recommendations.
Because they like data as much as they do, these set of people are domain experts who usually have a working knowledge of data science and Machine Learning algorithms. They want to capitalize on their domain expertise and leverage data to build their own use cases.
All they need is a data science platform that gives them exploratory and processing skills and learn and help them adapt to modeling techniques and build solutions
We have written about the inevitable rise of the citizen data scientist as data embeds itself in the DNA of organizations globally. The Business User has a critical role to play in a data-powered organization. Why? Because for the organization to become data-driven every decision has to be backed by data. And it is the business user who drives this.
But most business users are not data experts, right? How can they play with the data to derive useful business insights from the data themselves? And if they could imagine the kind of business advantage that it would bring. Imagine your HR team playing with data to track, analyze and share candidate profiles to avoid bad hiring decisions. How about your marketing team becoming data-driven to customize marketing experiences and improve marketing outcomes? The uses cases there are many.
In fact, I feel every department, every business user, benefits from using data to drive their decisions. And clearly, every department now needs its users to become citizen data scientists by giving them the capability to summarize, process, and visualize data by independently building and taking control of the machine learning outcomes. All you need to do is enable them with a platform that helps them do so.
Given the growing deluge of data, the job of business analysts gets more complicated. Today your business analysts need creative analysis capabilities given the plethora of data at their disposal. Think about it. Your KYC is no longer about the data that your customer fills in. It’s got social, behavioral, sentiment, and all related customer data to consider as well.
Business analysts have to get more creative in the manner they play with data. They have to make the data at their disposal work harder and work better. They need to ably visualize this data or plot results easily and with greater dexterity. The business analysts of today have to take business analysis to the next level because today he/she has the data wealth to do so.
But can traditional BI tools enable this? Especially when we know that BI also needs the data science boost? Your analysts will need to leverage data science, employ machine learning algorithms, automate the arduous math and coding components of data science to deliver dynamic visualizations that drive business growth.
Quite clearly, you cannot call yourself a data-powered organization if you are solely piggy-backing on a few exclusive data scientists. Just like how we have been able to democratize data, we also need to democratize data science. And this can only happen if we give our users the right platform and tools that can help them become citizen data scientists and make the use of data more deliberate.