Business decision making is becoming more and more ad-hoc. It’s a daily necessity. Even the best managers often prefer to trust gut instincts more than hard facts. This is partially due to old habits refusing to die, but alsobecause there are very real difficulties in getting the right information in the way managers want and at the speed they need. Faced with the rock of decision pressure and the hard place of inadequate data, they are left with little choice but to guess and hope for the best. The problem exists at every level. Everything from strategic planning to daily operations is taking a more ad-hoc flavor. Let me use banking as an industry example to illustrate the case in point. What follows is based upon years of experience in seeing how banks are often forced to operate.
Banks face enormous strategic problems in a world that swings between more and less regulation. Managing this environment is especially difficult in fast developing markets like the BRIC countries where the regulatory framework is often unfamiliar to that which exists in more developed countries.
There are armies of analysts at big banks whose job is to analyze huge amounts of data in an effort to discover trends and then pass that information back to the business to come up with better offerings for loans, credit cards, mortgages etc. Some of the scoring mechanisms they use have thousands of parameters applied across millions or even billions of records.
If the models that help decision making are too strict, banks minimize their risk but won’t optimize their use of capital because they have to turn away too many customers, who in turn will likely take their business to a competitor. If the models are too lax, then the bank could very well end up making higher risk lending decisions that in turn generate poor returns or lead to unacceptable levels of bad debt. Either way, banks are struggling to optimize capital while fulfilling their job of lending and deposit taking.
Opportunities are lost when the analytical models run in long batches. This is because there is a limit to how many times you can fine tune a model before it is time to go to market. Getting it right or at least ‘more right than other banks’ is key to a sustainable business model. Now let us consider problems at a more tactical level.
I like to take India as an example as it is a country with which I am familiar. The vast majority of India’s population are not serviced by any bank. For many people, the nearest ATM might be 20 kilometers or more away from where they live and work making traditional banking a practical impossibility. India is also a huge consumer economy that has thriving retail markets but where many businesses are in dire need of working capital. You can imagine how many banks want a part of the action. The key to winning that business is how well they can tackle this very real ‘big data’ problem, how to best target the ‘unbanked’ market with offerings, how to make quick decisions on issuing credit lines without losing the customer to the bank next door, what are the best locations to build ATMs etc. Incidentally, there is a very real mobile problem to solve too, that goes hand in hand with the big data problem. India has a lar=
ge number of people who use mobile devices – especially SMS. Mobile needs to be the front end of choice for any customer facing solution there.
At a more mundane operational level many banks have ‘keeping the lights on’ problems to overcome. A simple example is the daily analysis of financial line items – which items are open, which are cleared etc. Important treasury decisions are based on such simple reports and every second lost in generating that report is a potential loss of opportunity to make the right investment decisions. Historically, these companies have developed best practices based on an understanding of trends rather than current real time data.
While that provides a more secure way of understanding risk, it cannot account for short term opportunities. It follows then that even a small improvement in the speed with which decisions can confidently be taken has significant impact on competitive differentiation.
Given the need for ad-hoc ‘turn on a dime’ decisions, it is important that managers have access to a solution that solves their data problems. The keypoint is that increasing the speed of existing systems on its own is not enough. Banks need a platform that addresses important considerations. Examples include:
- Ability to massively scale. Data volumes never shrink, they always grow.
- Ability to work well with all kinds of data sources. It is becoming clear that the ability to consume bot structured and unstructured data is essential to a proper understanding of market conditions.
- Ability to work in harmony with complimentary SQL and NoSQL solutions
- Ability to use complex predictive logic. Data on its own is not enough.
- Ability for users to ask any number of arbitrary questions without a lot of IT involvement in back end. Humans have this pesky ability to dream up questions that still need answers but in today’s world there is little time to wait for others to come up with the answers.
- Ability to provide output in a way that any system or device can access.
- Reduce the number of hoops information needs to jump through (for aggregation, transformation etc.) before reaching the user can make decisions.
While we’re at it, there is an expectation that all of this needs to be done in a cost effective way, with rock solid security.
The reality is that no one company – including SAP, other vendors and customers – can provide all the applications needed to address these requirements. The scope is too broad. Does that mean we’re stuck? While there are various potential approaches to achieving the right solution, we believe the starting point is a platform that meets the main criteria established above, so that all the required applications can be built on it. However, any platform must bring with it the ability to deliver a demonstrable return on in=
As the leading provider of business solutions to global companies, SAP has already invested heavily in developing a solution for these big problems – no points for guessing, that is indeed SAP HANA. It is very easy to get started and developers can get free access.There is plenty of formal (via SAP Education) and informal Hana training (via Hana Academy , SAP Community Network etc.) You can even spin up your own Hana One instance on AWS marketplace. So don’t wait on the sidelines – jump in and check out the Hana experience for yourself.
If you want to see sample use cases of how Hana solves real business problems – find them here. And of course the proof of the pudding is in the eating – hundreds of our customers and partners have bought it and are using it to solve the big business problems that they could not solve before. You can find some of their success stories here. And of course you can expect to see the latest and greatest about Hana next week at SAPPHIRENOW in Orlando.
In subsequent posts, we will discuss specific details about how our customers and partners are making use of HANA, where we are headed with the platform and which frontiers beyond our traditional enterprise market we are focusing on.
Image rights via SAP