Big Data and analytics have generated lots of revenue for hardware suppliers, software providers and consultants. They have also created lots of jobs for people with skills ranging from basic statistics to advanced mathematical modelling skills. What is highly questionable is whether all this expenditure has generated value for the banks that have invested in them?
Like many new business philosophies and technologies the approach banks have taken to adopting them is to build them in-house. Just like when computers first emerged and individual departments took it on themselves to buy their own computer, hire their own programmers and write their own code to address their department’s specific requirements (as an aside, It is one of the reasons why so many banks still today have such dysfunctional IT departments and systems), banks don’t appear to have learnt the lessons from the past and are adopting the same approach when it comes to data and analytics. Functions such as risk, mortgage underwriting, card product management, marketing, finance and treasury are creating their own local data marts, hiring their own data scientists and modellers and buying their own query and advanced analytics tools. They are building models, sometimes in inappropriate tools, with inadequate testing that the bank’s executives are making critical decisions based on the output from them.
The fact that individual departments are doing their own thing is very cost inefficient is the least of the problems with this approach. Even for banks that have elected to go for a Centre of Excellence operating model for data and analytics whereby a central pool of data and analytics experts provide services to whole bank there is a fundamental problem with this way of addressing data and analytics.
Building models in-house is predicated on the basis that every bank is so unique that the models will provide differentiation from the competition. However banking, and particularly retail banking, is based around standardised products with standardised ways of underwriting those products, standardised ways of funding the products and very largely standardised way of moving the customer’s money. Therefore spending large amounts of money hiring expensive data scientists and modellers and then lots of time building models when there are standard models available to either buy or pay for the use of from the likes of Experian, SAS and other data and analytics specialty firms makes no sense. Not least of all because true data scientists need to be continually fed interesting and challenging problems to crack (something few banks will be able to consistently provide enough of while specialty firms will be able to) otherwise they get bored and stressed – a bit like caged lions that are fed raw meat rather than having the excitement of the hunt.
Unfortunately the peddlers of Big Data and analytics solutions don’t point out to their customers and the IT users who buy their solutions don’t acknowledge the critical fact that:
Data and analytics without context and insight is of no value to a bank.
Insight is an unfortunate word because many banks take it to mean having a better understanding of what is going on inside their banks. However that is only the half of it. As critical is to have an understanding and the context of what is going on in the environment that the bank is operating within. What are the competitors doing, what is happening and could happen in the macro economic environment and how would that impact the bank’s customers are just some of the potential questions that need to be answered to create insight. If there had been a better understanding of some of these questions then it is possible that the financial crisis of 2008 could have been avoided.
However insight of its own is not enough. A number of banks across the globe could rightly claim that they have teams of data scientists who like the PreCogs in the Tom Cruise film ‘Minority Report’, who were able to predict crimes before they were carried out, know so much about their customers that they can predict what they will do next. However having that knowledge but not having a means of sharing it in a simple and usable way with the banks’ systems and the people who use those systems means that it is of no value at all.
Insight with the ability to know the ‘Next Best Action’ and execute on it is what will define the banks that will emerge as the winners.
This ability to apply data and insights to bring about great outcomes should not be limited to use with customers but should also be applied to other areas of the bank such as pricing models to allow personalised offers, to fraud detection, to identify money laundering activities and to make better funding decisions. The list of areas where this could be applied to banks is almost limitless.
However what it requires is a very different approach to data and analytics then is largely adopted today. It needs to be driven down by the desired business outcome with the data required being seen as the very last thing. It needs to be driven by the business executives not from IT or worst still technology vendors. It needs to be driven as an industrial process rather than as a cottage industry. Banks need to understand that where they will be able to differentiate themselves from their competitors is on their insights and how well they execute on those. For the rest they should look for best in class products and services for data and analytics from organisations that are truly expert in those areas.