Analytics in the World of Business — Here’s what I learnt from Kellogg
*Everything written here is based on my key takeaways from the Executive Education Program I recently attended at Kelloggs, North Western University.
“Without big data analytics, companies are blind and deaf, wandering out onto the Web like deer on a freeway.” -Geoffrey Moore
Inspite of so much data available to companies these days(in some cases the data is readily available, and in others, can be collected with some effort), it can be argued that most companies aren’t doing enough to tap the value that dataand analytics can bring to the table, in terms of aiding management in both strategizing and execution.Analytics, at its best, can be leveraged to create a strong competitive advantage and build an exceptional customer experience.
The popular case study of Paul DePodesta of Oakland A’s, portrayed by Brad Pitt in the movie Moneyball, was how analytics becamea competitive advantage, and how it helped a team with very limited resources to win consistently. This is from 2002, where it was believed that the adoption rate of analytics would sky rocket overnight- but contrary to what was a common perception then,the actual adoption rate has been slow and gradual over all these years. This is due to a variety of factors.
The depth of the usage of analytics is not a question of technology or statistical techniques. Analytics is fundamentally a leadership and business problem. It requires leaders and managers to articulate the problems and opportunities in the right manner, and encourage the capture of data and usage of the right analytical techniques to help them in their decision making (Note the emphasis on “help them”. Analytics cannot make the decisions for you, rather it only aids in the same).Once the data is thoroughly sorted and analysed, it is then up to the manager to make the right decisions using his own judgement.
Let us take Customer Analytics as anexample. Before thinking about the proper usage of data, it is important to understand the underlying data in depth. It is very important for the managers to make sure that their systems capture all the relevant data(structured and unstructured) and processes it in the right manner. A common mistake that companies make these days is resorting to only the traditional sources of data. Whereas, when looked around, we are surrounded by potential data, for instance, online customer sentiment. Data should be channelized and used properly from all sources such as the Transactional and Application data, machine data, data from social networking sites, and so on.
Once it is done, it is important to use the right techniques to solve for business problems. Here, problem articulation is critical. It need not be just business problems, but also can be directed to uncover untapped opportunities. Relevant techniques are then used to uncover insights.Based on the problem articulation, one needs to decide which analytical technique (or a combination of techniques)is best suited for this. There are three broad approaches- Descriptive, Predictive and Prescriptive analytics. Where each of these can be used separately, it can be used in a synchronised fashion too.
There is one common perception that they are a progression, Prescriptive is more evolved than predictive and so on. This is incorrect, as the approach depends on the problem to solve for. There are many examples of good descriptive analytics used to solve critical problems.
Descriptive Analytics looks into the past. It is used to report, visualize and understand the past data. There are great visualisation tools like Tableau that can be used for effective management reporting and decision making.
Predictive Analytics is about anticipating outcomes by understanding the underlying relationship between data inputs and output. For example, if the business problem is high churn of customers, a modelling can be done on various parameters to predict likely churn of a customer and appropriate campaigns run on a regular basis. And Prescriptive Analytics on the other hand determines which decision produces the most effective solution. These approaches do not rely just on traditional techniques such as the regression model but also uses powerful modern techniques such as machine learning. It is up to the managers to use these tools in perfect synchronization to achieve the full potential of Analytics.
While all these analytic tools are about the present form of business, the future is all about experimentation. Use of techniques such as A/B testing (a form of experimentation technique upcoming in the business environment today) can help the manager in deploying best customer solutions in the market. And while its usage is very high among digital giants like Amazon and Google, old world companies would also find high applications of experimentation in non-digital contexts. It can influence decisions encompassing pricing, sales incentives, media inputs, etc. A very simple approach to experimentation could be testing two different sales incentives programs. For this, one could do it in two different teams/ geographies, compare the results (A/B comparison) and decide which one is better.
Data also creates new paths for understanding customers and enhancing the customer experience.Analytics available today provide companies with an opportunity to create business models to drive their day to day decisions. In India, some BFSI and Digital companies have already leapfrogged ahead of others by adopting analytics. Firms need to pilot and develop their journey in this much untapped space of analytics. Companies that can take the best advantage out of data analytics today will set themselves apart and have a large competitive advantage tomorrow.