SAS Says Speed Key to Marketing and Fraud Detection Analytics
While all but the most clueless companies know about the growing silos of unstructured data that are piling up in organizations, relatively few have figured out what they should be doing about it.
“Most organizations know they could be extracting far more value from the various data sources they hold. They are struggling to get to grips with the basic elements of how they can analyze what they have when they know they probably don’t have enough skilled human resources,” said Tony Lock, an analyst at Freeform Dynamics.
Business intelligence (BI) vendor SAS is gearing up to address this issue with a series of recent product announcements that are tied together within its growing analytics portfolio. Two of these offerings are tied specifically to the marketing and financial fields, both areas that analysts like Lock agree are ripe with potential for leveraging insights gained from analyzing so-called Big Data.
SAS High-Performance Marketing Optimization has been designed to analyze millions of rows of customer data to solve complex marketing problems, enforce contact policies and stay on budget. For example, a marketing unit operating many different campaigns can run what-if analyses to predict the impact of changes.
This SAS marketing tool integrates with SAS High-Performance Analytics, which provides an in-memory architecture to accelerate BI analysis. Instead of one or two processors having to crunch all the numbers, SAS has adopted a massively parallel processing (MPP) approach which uses multiple blade servers in parallel, each containing multiple processors. The result is said to shrink compute time from days to minutes, even if tens of millions of customer records and variables are involved.
“SAS High-Performance Marketing runs on a high-performance architecture where the data and processing is distributed,” said SAS vice president of sales development and project management Randy Guard. “You can quickly look at campaigns, budgetary or regulatory constraints, and set your goals which are then run through an optimization engine.”
Making Marketing More Agile
Speaking at the recent Premier Business Leadership event in Las Vegas, he offered an example of a company with 15 to 20 million prospects and eight separate marketing programs. Each prospect could potentially receive as many as 20 communications per day based on the ongoing initiatives. The challenge is picking the right communications for each person for specific days – a task that can be tricky, given that companies want to inform them without overloading them.
“That kind of analysis took one customer up to six hours before, so they could run it once a day,” said Guard. “Now, it takes 1 minute and 40 seconds, so they are able to change their direction during the course of the day based on immediate feedback.”
Take the case of a product launch. One hour into it, the analytics app could detect rising negative sentiment about one aspect of the program. The software allows managers to quantify that sentiment and adjust their campaign, if necessary. Another example: If a competitor counters your offer with a better one, you can model the consequences of changing your own offer and determine what to do.
In a similar vein, SAS Revenue Management and Price Optimization Analytics addresses the pricing problems of the hospitality, travel, sports and entertainment verticals. It helps companies interpret how competitors are pricing items and how that might be impact current demand. MGM Resorts deployed this approach to improve its pricing decision-making and free up managers from extracting and manipulating data.
Fighting Fraud
A major revelation at the SAS show was the growing importance of analytics in improving fraud detection. “Fraud management is the single biggest area of growth for SAS,” said Guard. “The challenge has been to minimize the false positives.”
Accordingly, the company has released a tool aimed at detecting money laundering and other financial crimes. The SAS Financial Crimes Suite consists of several modules which can be deployed singly or in combination. Components include data management, detection and alert generation, alert and case management, predictive alert analytics and fraud detection, as well as dashboard reporting.
Analytics and Need for Speed
So critical has become speed of analysis that SAS is even dispensing with the need to collect it first. Known as the analysis of events in motion, this technology takes data and analyzes it as it is received. This is done by the SAS DataFlux Event Stream Processing Engine, which can look at both structured and unstructured data.
“Generally available in December, this event stream processing engine will be able to capture and analyze data in flight,” said Guard. “Depending on the content, the company can decide whether to then store it or not. This tool can be valuable to risk managers, for example.”
Guard made much of the SAS MPP architecture. Instead of trying to analyze data on disk, the new version of SAS analytics places the data in memory and throws far more processing power at the problem. Result: conclusions reached two or more orders of magnitude faster.
But Freeform Dynamics' Lock believes that getting BI users to consider the relative merits of in-memory analytics and MPP may not be an easy task. As with any new approach, enhancing awareness and educating users will be a significant part of selling the solutions. After all, the business intelligence community tends to focus on software, not hardware.
“SAS and the other providers of BI tools would do well to help those unfamiliar with modern solutions to understand just what is available and how they can get started,” Lock said. “Only after these steps are taken will they be ready to look at more sophisticated offerings as they start to see value from initial investments.”
Drew Robb is a freelance writer specializing in technology and engineering. Currently living in California, he is originally from Scotland, where he received a degree in geology and geography from the University of Strathclyde. He is the author of Server Disk Management in a Windows Environment (CRC Press).
Drew Robb is a writer who has been writing about IT, engineering, and other topics. Originating from Scotland, he currently resides in Florida. Highly skilled in rapid prototyping innovative and reliable systems. He has been an editor and professional writer full-time for more than 20 years. He works as a freelancer at Enterprise Apps Today, CIO Insight and other IT publications. He is also an editor-in chief of an international engineering journal. He enjoys solving data problems and learning abstractions that will allow for better infrastructure.
