Various ways how Data Science and Machine Learning influence or impact Business

Recently, with an intention to explore the recent trends, market position of various technologies and organizations with respect to about Data Science and Machine learning, I was surfing Gartner surveys and various articles. Few areas attracted me and have increased my interest to go into more deep. One of those is various ways how Data Science and Machine Learning influence or impact business. It is an area which is becoming more and more critical now a days to even drive business. It is not new that analytics plays an important role in business by helping CxO to take any business decision, sometimes important decision to even disrupt market. Recently, I also read an article of Bernard Marr where he compared Data with Oil and concluded that Data is not new oil. However, I believe that it is not good to compare data with oil literally as data is not oil. I think it is on top of some other article where someone said that data is data is new oil for business. I interpret it as new opportunity for business and it is like new fuel of business. In many other article, Bernard also highlighted the same. Bottom line is data is important and non-separable part of business.

Anyway, till date, traditional BI, self-service analysis are the tools or approach are fulfilling it. However, traditional approach and technology have many limitation in terms of data and finding insight. As we have various new technology and approaches now a days in Data Science and Machine Learning area, it becomes now important for business to be dependent on this. Due to this fact, most of the organizations is now building up new team and mostly called as data science team. Therefore, the team who is responsible to implement this area within the organization has become critical and their responsibility also has increased. Hence, it is important for them to focus on few areas which can impact or influence business. I like to share the summary of my reading.

1)      Innovation – New thinking to disrupt business through data science

With data, traditionally we are already doing analysis and coming up with solution. However, to do so, we are applying traditional methods. Those methods are having inherent limitation to handle complex problem. Due to lack of technology, we are compromising with such solution. However, with the advancement of methods and technology it is now possible to handle complex problem and even new way of deriving solution of old problem. Therefore, it is important for the responsible team to think and apply innovation – important to attack any problem (old or new) with new thought by going beyond traditional thinking.

 2)     Exploration – To explore new and uncovered pattern hidden in data

Due to technology and cost, till data we had one limitation to handle large volume and variety of data. We gradually have overcome this limitation with the invention of big data technology, NoSQL etc. It opened up an opportunity to work with huge data – that can reveal various dimension of insight. Hence, it is also important for team to explore raw data in every aspect to uncover various hidden pattern by applying out of box thinking even there is no clearly defined objective. New pattern can then set objective for business.  

 3)     Prototyping – challenge with new solution

While data become huge and complex, then manual analysis and decision making ability of human being are not sufficient to drive business for new dimension. Therefore, technology becomes efficient tool which can complement the decision making capability of human. Data science with AI/ML methods sometimes can reveals surprizing solution.

 4)     Refinement – Continuous improvement of existing solution

This is very important factor. Data are different in nature and there is no uniform single rule which is always better for all kinds of data. Every kinds of data has its own nature and therefore, it is important to have data exploration as explained earlier. If a solution is modelled and made live is not lifelong solution. Over the time, it is inevitable that input data will change due to change of business, addition of new dimension of business, change of environmental parameters and so many factors like this. Therefore, it is very important to have a process to monitor existing model and continuous refinement to make it always relevant to current situation to bring correct value to business. Otherwise, running solution might impact business adversely.

 5)     Firefighting – Identifying the driving factor of upcoming situation

In many cases, organization needs cause of certain critical problem in crises situation. In those cases, it is important for data science team to identify only the cause of problem – not like data exploration. True that it is also one kind of data exploration, but application is different. Once can argue that for this, traditional analysis and self-service BI are sufficient, but if we have advance technology, methods of data science available, then why not to take help of that. It might come up with new interesting dimension of cause of those critical problem.

 There are many other factors which can be listed. However, I believe these five areas are most important. With this, organization, can have better understanding of any burning problem even at crises situation, better understanding of customer behaviours to provide better customer experience, reduce customer churn which is nothing but protect business loss, better way of driving business with new innovative initiative to increase business. Therefore, just building data science or any special team is not sufficient, it is also important to have proper strategy and goal for that team – which then can add value to business properly.

It’s summary of my thought which I have borrowed from those Gartner articles. However, I’m sure there are many other viewpoints. Hence, I would also love to hear from all of you if you share your view in this regards.