Barriers to Data Mindset

 It is very hard for a bank to develop a data mindset.  Traditionally, regulated banks have been trained to deliver accuracy as they are managing money. The focus is on getting it right without errors and making sure all compliance requirements are met. This culture does not reward any creative inputs. Legacy systems on which banks have spent decades managing are also inflexible. In addition the process of completing a transaction has to follow deeply structured processes.

This environment does not allow empowerment of any sort at the customer level. The demographics of the customers using the banking services have changed. They are younger, often more affluent and tech savvy . They are used to Amazon type services and Apple type user experiences. If a bank cant provide what they want, they will look to fintechs who can. Post pandemic this customer churn is hitting 15% .

So what can banks do? Its well established that a customer centric service paying attention to unique situation a customer is in their financial journey can make them happy campers. This level of service can be obtained by gathering and harvesting insights from the data the bank has about the customer. Banks have not found any meaningful ROI with their investments in data technologies. They still operate in silos so much so that a late 2017 study found they are valued at 54% below book value.

The answer lies in a top driven change management approach.

First, get the KPIs of the top level executives to include metrics on data initiatives, these should be tied to their bonuses. At the second tier of executives empower them resources on obtaining education on how data driven organizations work. At the third level get together small cross functional teams with clear risk and reward for targets identified to use data and improve operational processes, reduce time spent by customer on interaction with various channels to name a few. At the front line level make available the complete profile of the customer and recommendations on talking points when interacting with customer. Reward the frontline for usage and success stories making them stakeholders in the data mindset culture.

These are not impossible actions. Consider what a 5% customer retention means, a 15% uptick in new account openings, a 10% increase in loan book means to the bottom line of the bank and see the trade off between status quo and investing for a data mind culture in the bank

Read More

What is likely to happen? The journey of Prescriptive Analytics

 This is the final frontier in data analytics-- the ability to respond in real time to event data in a way that allows you to back the instincts of the decision maker with data. If your data is telling you what your instincts are saying, the chances of making an informed decision to influence outcomes increases.

However, the right questions have to asked of the data otherwise the outcome will not be what was desired.

Some examples in banks where this type of data analysis is prevalent:

Customer insights based on NLP based sentiment analysis which can be fed into the mobile app

Trading intelligence on what stocks to buy

Better credit scoring by using non traditional sources like social likes .

Read More

What can happen? The story of Predictive Analytics

Predictive Analytics is the process of using a combination of data analysis, statistics and machine learning to make predictions on possible outcomes.

Examples of these can be:

-predicting when a turbine in a power plant fail operating --this can help schedule maintenance activities  

-predicting customer behaviour based on time of our marketing promotions

-predicting capital required to meet regulatory requirements based on the volume of loans a bank has given in a particular period of time.

We use well established statistical tools in machine learning also known as supervised machine learning where we are trying to arrive at a number to predict the outcome. The fact that you are able to use a data set to run an ML algorithm means you are well advanced in the journey of analytics. It assumes you have the sources of data in the format you need and they are able to work inside a model developed for the purpose of prediction of outcome.

Read More

Why did it happen? The Story of Diagnostic Analytics

 Diagnostic Analytics tries to answer the why something happened in the past  from our data sets- for example why did sales rise this month though we didn't have any marketing campaigns, why did patients in a sample have the same symptoms,why did our bank's customers choose the 5 year fixed mortgage product.

Frequently this side of data analytics looks at both internal and external data sets to draw inferences . The process of drawing out the inferences can lead to a significant benefit-- the creation of a data mindset in the company that will have force multiplier effects on the company's future performance . The process itself leads to exposure to tools for data drill down, statistical methods and confidence in future decision making --all leading to a virtuous cycle of creating a data mindset.

Read More

What happened? The story of Descriptive Analytics

 The oldest know data analysis consists of two things-aggregating or summarizing data and  data mining.

The majority of a company's investment in data analytics is focused on aggregating data-- what was the sales, by month, by region for example.Comparing to previous time periods gives an understanding of where the company is going for instance the sales have increase 20% compared to previous quarter. It is an important tool for analysis of the performance of a company. Different types of data are then mixed and matched to provide further insights such as :

ROIC--Return on Invested Capital takes  three datasets--net income,dividends and capital .This allows us to compare performance of two or more companies.

On a similar vein, company annual reports are descriptive analytics in action . They tell the story of what happened in the past. These reports are mostly derived from spreadsheets and ancillary sources such as ERP tables .

Data mining techniques allow you run the group aggregated data and look for logical conclusions to understand customer behaviour. A grocery chain may use data from its loyalty card program group to analyze what are some of the fast moving items in the produce section.

Banks are one of the best examples of the descriptive analytics mindset. They are so focussed on the historical aspects which is important since they are custodians of our money. The problem is both legacy systems and the lack of new data mindset is making it very hard for them adopt the newer types of data analytics 

Read More