Expert Interview with Cal Braunstein from RFGonline.com on Big Data Analytics
How can data analytics help businesses get on top of risk management?
An average large company has more than a dozen regulations that they need to conform to. Failure to comply results in a risk exposure which could mean costly penalties, fines, or worse, data breaches. Moreover, each of these regulations addresses more than 100 control domains, which can break down to thousands of policy specifications that must be complied with. That is tough enough, but all of this must apply to all applications, systems, infrastructure, and processes (IT and non-IT).
Thus, for any single day there could be 100,000 elements to look at. This gets even more complicated in that for each element we are not talking about a point in time, but a span of 24 hours (or for however many hours a day a firm operates). There is no way an individual can easily pore through this data and spot problems.
Through the use of data analytics – historical, realtime and predictive – a company can spot current risk exposures or the trends that indicate an upcoming exposure. Early recognition of risk translates into an early response; which, in turn, reduces the magnitude of the exposure as well as the risk.
How has data analytics helped improve business processes up to this time?
Traditionally, data analytics was historical; i.e., it looked at past transactions and showed trends or performance metrics of some kind. So for example, executives could know who were the best performing sales teams or regions, or best sellers, or products with best margins, etc. A company could learn the value of each customer or what customers buy individually or in combination. These data points would help in developing sales plans, product selection/improvements, site selection, customer retention approaches (Know Your Customer information), etc.
A great example of this is Progressive Insurance. They have all the motor vehicle information on individuals in a huge database. They know in advance the risk associated with each person who wants to apply for insurance. For individuals that are low risk, Progressive offers a good deal and tries to win the individual’s business. For drivers that are a poor risk, Progressive is happy to show that individual that he/she can get a better insurance deal from a competitor. Progressive wins; the competitor gets a lower margin driver.
Then over the last 15 years, analytics has been improved so that it can be used in real time fraud identification, and in up-selling or cross-selling customers while on the phone, in the store or on the Web. This reduces shrinkage and bad debt loss and/or drives up margins and the per-sale value of each customer transaction.
How do you see data analytics contributing to the improvement of business processes heading into the future?
Data analytics is now moving into predictive and preventative analytics, which are two new exciting areas.
Let’s examine that from a healthcare viewpoint. If we know enough about a patient, it may be possible to predict the next set of events and potentially act so that it is prevented from occurring. One big example of that these days is in diabetes care. Insurance providers are checking on the indicators of pre-diabetes or diabetes in individual customers to determine the potential for the individual to become diabetic. In those where the indications of diabetes are high, the companies are proactively reaching out to the individuals and attempting to change habits or patterns so that the probability of becoming diabetic is greatly lessened. The insurance companies recognize that it is more economical for them to work with clients to reduce the number that will become diabetic than it is to cover the cost of diabetic care for the individuals over a lifetime.
How much has data analytics affected procurement and the supply chain so far?
Procurement has not been a major area of use for data analytics, but it should be. Companies spend a tremendous amount of their revenues on the acquisition of products and services – in the purchase of or re-purchase of these items. Yet in most cases, companies lack the information needed to get a better deal initially or thereafter.
For example, a company acquires 10,000 PCs from a vendor over a 3 -year period and will replace these boxes every 3 years. How good were the PCs? What was the failure rate for each type of PC? Which ones had high maintenance? How many came in “dead on arrival?” Before going back to the same vendor and re-ups for another round, the buyer should know what he got and use that information for negotiating a better deal next time.
In the manufacturing industry, where a single product such as a nail or screw or bolt is used hundreds or thousands of times on a single product (such as a car or plane), it is important to know the failure rate. The cost of using the right nail made to the proper specifications is far cheaper than paying for repairs when the device breaks down. The company will only know which products from which suppliers are meeting specifications if the data is captured and analyzed. While a screw is very inexpensive, when multiplied by the thousands it becomes real money. Negotiation counts. And when it fails and must be fixed, it can be 10x to 100x more expensive than the initial cost associated with the original build.
How do you see data analytics affecting procurement heading into the future?
Procurement departments can use it to determine who are the best suppliers that should become strategic suppliers, as well as to weed out the poor performers. Hence, companies could shrink the number of suppliers they use and thereby cut purchase costs (through better information-based negotiations) as well as the internal expense associated with acquisitions. With the right data procurement, departments can better negotiate pricing, discounts, and volume purchase agreements. Also they would be able to know more about shrinkage and spoilage, which in turn can be translated into better products or less waste.
Additionally, analytics can be used to ensure the buyers are complying with the purchase rules and not acquiring goods “off book” from non-preferred sources which tend to be more expensive. Lastly analytics can be used to move toward just in time (JIT) purchasing. This eliminates overstocking and wasted resources (including funds).