In my last blog, I mentioned the massive revenue opportunity for monetizing data after signature of the Open Data Executive Order. Similar to the open commercial access of GPS satellites, we’re poised for a big data breakthrough as 10,000 new sources of data become public and machine readable. Now, where do we put that data and how do we work with it?
Myth #1 - ERP is not big data.
CIOs have started big data initiatives from ERP systems, and I don’t blame them. They have reputations, investments, and thousands of transactions from call centers, suppliers, and customers tied up in ERP. The numbers will show that ERP is just a fraction of the data that exists about customers, their location, your engagement, and your markets. While ERP vendors offer the myth of big data, based upon customer details and transactions, they inherently lack big data volume and variety.
Sensors in cars, mobile device activity, traffic patterns, census data, and new housing permits and thousands of other data varieties tell us stories about clients far removed from ERP systems. Furthermore, ERP systems will offer the basic reporting tools to inform organization how many of SKU# 123 were purchased, when, and by whom. Unfortunately, you’ll never know much of the why to become predictive and self-aware. This is were ERP, and their in-memory capability, falls short.
In a simplistic example, companies that sell building supplies tap into ERP data to know what products are being purchased by builders. The companies, and their ERP systems, are unlikely to incorporate census data, home permits issued, employment trends, and hurricane predictions to effectively predict time, place and quantity of inventory. Property insurance companies provide excellent examples whereby they integrate customer data, geography, replacements costs and weather histories for hurricane models when they decide to exit select states, raise premiums, or decline new business in coastal markets.
Volume and variety of data resides outside of ERP while users and data scientists require holistic data access for decision making. ERP vendors suggest organizations converge OLTP + OLAP into one common memory, creating unpredictable license costs and platform performance risk. If you are serious about monetizing data, and knowing everything about your clients, there is an alternative. IBM Business Intelligence Pattern with BLU Acceleration is an in-memory solution built upon a PureApplication System appliance, set apart from single-stack and static ERP platforms. Considering the limited data captured from ERP transactions, organizations should remember big data is the strategic differentiator, not the ERP vendor.