As a consumer, I recently went through an experience that led me to write on this topic.
I was looking to buy a new car. I browsed for reviews on Google and watched videos on YouTube to get deeper insights into data about the make of different cars. I was aware that my activity would get tracked, and I would receive targeted ads. I was surprised by the relevance of these targeted advertisements that I received online. In the first week, I got advertisements for cars in a specific segment of interest. In the subsequent weeks, the algorithm figured out that I am a brand-loyal consumer, and I received the content from only one car manufacturer. When I zeroed in on the model, I started receiving content about the benefits of automatic transmission of that model. However, the barrage continued for a week after my purchase. Perhaps the fact that I did not post a picture of the delivery of the car on my social network is responsible! Algorithms must get better at personalisation.
Each time we make a payment or ‘like’ a picture on social media or browse information about any subject, our activity is tracked with our permission when we use these online platforms. This data is then analysed to find meaningful patterns. For example, the type of products that we buy online using our credit card can throw light on disposable income. Our social media interactions or posts can tell us something about our social status. Our online browsing provides information regarding potential purchases. Data Science techniques are then used to connect these dots and create targeted digital marketing content for us. If you pay close attention to your buying experiences, it is easy to figure out that data science is not hype but a real shift that is affecting us in our daily lives.
Traditionally, data science was focused on statistics, but with the advances in technology, algorithms are becoming smarter at reading images and text. Today, data science techniques are making their presence felt in almost everything integral to our lives, right from predicting weather to forecasting our buying behaviour. If we review our entertainment experience on television, we can see that machines have started influencing, or, to some extent, controlling our choices on platforms like Netflix or Amazon Prime. Algorithms are also predicting our online shopping behaviour and influencing our decisions.
Algorithmic trading is a reality of life. Two critical shifts will shape the future. The first one is the emergence of new technologies that recognise a human sense like voice, eyesight or touch. The second one is, machines mimicking human senses like vision, sound, etc. The combined impact of the two shifts can create new data streams and possibilities. It will not only have an impact on the way we do business but also the way we live our lives. We are facing a future where it might get difficult to differentiate a human being from a human hologram created by machines. What does this mean to us as data science professionals? First, it means a huge opportunity to participate in shaping the future. At the same time, it creates new industries and career avenues with roles in data science skills continuing to dominate the job market in the current age, and for several years to come.
Today, we look at data science as one career stream. With the advances in technology, new specialisations will emerge. Let’s assume that you want to set up a data science practice for an on-demand entertainment video services company. An important starting point would be to devise the data science strategy. It will evaluate possibilities with various data streams, define the winning aspirations or goals and then create the implementation plan. Next would be to design the data architecture, which will ensure efficient storage, analysis and usage of consumer data generated through multiple data streams. These data streams include consumer ratings, voice, as well as consumer feedback about the service. It will be followed by the requirement to apply prescriptive data science techniques.
As of today, most of the data science usage is centred on descriptive, diagnostic or predictive analytics. In the future, the new-age data science practice will allow the service provider to generate content that is profitable and enriching for the consumer. Let me elaborate on this further. In one household, there are different consumer needs for online content on platforms like Netflix or Amazon Prime. My content consumption as a business professional is very different from that of my teenage kids.
Today, it is difficult to track the individual user preferences as the service provider might not understand the actual user who is holding the remote in his or her hand. However, once we move to use voice, it will be easy for the machine to understand if the consumer is an adult or a teenage kid. Within a single user ID, then, the content that will be pushed will be very different and more relevant for the consumer. Once, such interactions start between the human consumer and the machine that understands the human voice (tone to predict mood/emotions), there are limitless possibilities to personalise the content, and then charge a premium for it. There is another side to this argument. Consumers are apprehensive about losing the freedom to choose. In my opinion, we are smart enough to use these machines to our advantage. Only time will reveal the real answer to this question.
To conclude, I see the need for interdisciplinary skills to fulfil the needs of emerging disciplines in data science. Those adept in data management and storage will be vital for roles like data architect. Data science management will need senior management or leadership talent trained in it to manage data strategy.
