Predicting the Future Out of a Pile of Data
So I’ve been thinking lately about how organizations find their way out of huge volumes of data. I guess the techies have figured out how to make data from the past understandable with all their whiz-bang data warehousing tricks, but we all know that is like driving forward while watching out the proverbial rear view window. How on earth do leaders make fact based decisions that help improve their businesses going forward? In essence, how do they predict the future?
We all go shopping, so using a retail example should make sense to you. Think about the number of people that make purchases through a store’s single cash register all day long. Then multiply that by all the cash registers of a retailer. Then multiply that by all their stores … in every city, region, or country … throughout the entire day, 365 days a year (less if you live in countries that actually take holidays off to enjoy with families). That’s a lot of data! Now what would you do if your boss asked you to figure out, based on all that data, which products sell together, which of these products drives the purchase of the related products and what is the optimal price of each to maximize sales and profits? If you’re a retailer – I guess you would call it affinity analysis. If you’re not – well, it’s about figuring out which products sell together like chips and salsa, and then selling them at the prices that will produce the best impact to your business. This is a fairly obvious example to illustrate the point, but consider there are hundreds and thousands of far less obvious items purchased together. Understanding those relationships, concentrating on the most relevant relationships, and effectively pricing those items can create a significant competitive advantage.
Can it really be done? And let’s take it out of the theoretical … can it be done and actually produce optimal financial results? My colleague, Tony Collura, works in the group at SAP that is focused on advanced analytics and predictive solutions and he says “Yes it can!”
After that, he went all corporate on me and said “Our affinity analysis provides fast, directional insight from the granular relationships within the data to help develop strategies and actually forecast and assess future performance. This retail example is one of many problem spaces where our predictive solution(s) will help companies improve their competitive position with forward looking decision making; resulting in both strengthening customer loyalty and increased revenue and margin.“
Tony just wouldn’t stop. He started throwing out more examples of questions and challenges that our customers in various industries are constantly faced with:
l How does a retailer balance profitability with customer loyalty?
l How does a fashion retailer optimize depth and timing of markdowns to boost sales?
l How does a banker grow deposits without excessive interest costs?
l How does a manufacturer recognize if a new variant will cannibalize sales of existing packages or how the market will react to price changes on base models or extra equipment?
l How does a utility better predict load demand and improve customer satisfaction?
So it’s kind of like having your own crystal ball. Tony went on to tell me that business executives are also leveraging advanced model based decision support capabilities and tools across lots of industries to reliably forecast and predict future demand and efficiently assess alternative business strategies. This allows executives to set better strategies, understand their implications into the future and improve business performance. Is essence; extending their business beyond transactional efficiency and toward performance optimization. This evolves the organization’s activities from a sense-and-respond tendency to a predict-and-act. And to keep it real… You can watch a couple of examples of our predictive capabilities by clicking here and watching the demos in the “Optimizing with Analytics” section of the page.
Did I pique your interest? If yes, you can poke around on our web site for SAP Performance and Insight Optimization (predictive analytics). And if you don’t find anything that relates to your business, I’ll throw in a pitch for my line of business (that would be the SAP Custom Development organization) – we’re totally synched up with the analytics folks and are able to custom build really cool predictive analytical apps too.
And lastly … SAP HANA. Can’t write a blog without throwing that SAP buzz word in. Tony would kill me if I didn’t say that the speed of SAP HANA combined with predictive analytics is a dream come true for business leaders. Oh yeah, and it works on SAP NetWeaver and SAP BusinessObjects too.