We need more Data Humans

Data is Vector, not scalar.

Data has a character.

Data communicates.

Data reflects culture.

Data helps ‘drive’ decisions.

Data is not a Noun, it is a verb.

There is a human side of Data.

The fundamental shift with evolution of Data Science to Decision Science to Data Storytelling and Data Evangelisation is based on the premise of putting a value attribution towards data and analytics efforts. Despite great growths in overall adoption of data led strategies across industries , with significant investments, there are few organisations who have truly been able to integrate data into decisions. The challenge here is the gap between actual value chain vs perceived value chain that analytics can help build.

Organisations expect that data and models built on data should come up with a “Recommendation” that can either fix a problem or bring incremental dollars instantly. On the contrary, data is actually a process element. Predictive models are only good as long as you know the drivers and the levers that influence the future state, than just know what the future state would be.

“Accuracy” is over-rated. In most of the business scenarios, it is less about knowing a magic number, but the context around that number and ability to change something in order to influence that context. Data Accuracy and Veracity is an ever-existing problem to solve, as the data systems are not a static, rather always in flow.

“Speed of data” and further speed of analysis doesn’t match the speed of decisions in most of the organisations. Not every single time you are solving a herculean problem that needs to have the “best fit” model in place. Data Scientists need to become agile and integrate with the speed of business. Business decisions cannot only be data driven and the data science cannot only be based out of sanitised data. I have seen analysts being obsessed about data quality. Data preparation has been established to be taking most of the time from analytics practitioners but how good is a good data with a late appearance.

?Automation, Machine Learning and Platforms do not limit a data scientists ability or existence, but just improve the qualitative opportunity for connecting more with dimensionality of data, than size.

Data Preparations can be mechanised, Models preparation can be automated but the character of data needs human interactions.

We need more people who can sit between the “Outcomes of data Science” and “Impact of Data Science. These people need to be good with everything that is Not just science – Storytelling, Business Acumen, Consumer Psychology, People behaviour, Imagination and ability to connect the dots.

We need more data humans!

?