2a39 Oversight Systems | Blog

Big Data Advice to CIOs

After reading Tech Journal’s recent article “Big Data Comes with Big Obstacles for Some CIOs,” my advice to CIOs looking at Big Data as a challenge is to identify the keys to make the era that is upon us successful. One such key we’ve identified at Oversight is uniting multiple silos of data. And clearly one of the key silos is detailed information about your customers. The 2012 presidential election should provide the inspiration (or the fear if you don’t act) for gaining a deep understanding of your customers and prospects. The Obama campaign first united their own data sets on the electorate – they combined voter data with fundraising data as an example. They then incorporated external data sets.  Further they found ways to work around privacy concerns with certain data sets by “anonymizing” the personally identifiable information. In their case that let them gain a deep understanding of television viewing habits. The important concept is they found a way to utilize every data set possible.  And with this detailed picture they were able to target specific voters with the precise message most likely to convert votes for Obama. When an Obama campaign worker met with a voter they knew the pitch to use with that specific voter. When they sent a direct mail piece it contained a message meant for the individual addressee.

The power of Big Data is gaining the insights about specific events and opportunities and then to get those insights in the hands of the frontline.  Detailed information about your customers (and your prospects and your vendors) is a prerequisite.


 

Thoughts on the AGA Conference: Experts Agree “Pay and Chase” Not the Answer

AGA National Leadership ConferenceIn his opening address to the Association of Government Accountants (AGA) National Leadership Conference, Gene Dodaro, Comptroller General of the Government Accountability Office (GAO), mentioned “detect and prevent improper payments’ at the top of his list of how the accountability community can help to address fiscal and performance challenges facing Federal Government.  With Federal Government budgets already tighter and facing further belt-tightening and scrutiny with impending sequestration, Mr. Dodaro emphasized that the ultimate goal is to prevent improper payments from occurring in the first place.  He also mentioned that in 2012 four programs didn’t even provide data on improper payments and six others who did had their estimates and the accompanying data rejected because their estimates did not meet OMB guidelines for providing the estimates.

Great work is being done in a number of departments and agencies to achieve this objective, including some of Oversight’s most successful government implementations at the US Department of Defense (DOD) Defense Finance and Accounting Service (DFAS), the US Department of Education (ED), and the Bureau of the Census.  DFAS, ED, and Census along with other departments and agencies leverage continuous monitoring and analysis along with the automated insights this produces to predict improper payments that are going to take place so that steps can be taken to prevent the payments from occurring.  They also leverage these systems to address process or systemic issues to prevent the same thing from recurring.  Taking into account the cost of hardware, software, IT operations, and financial management operations, Oversight has seen cost to prevention ratios ranging from $3 to $30 of improper payments prevented for every $1 in monitoring and automated insights costs.

Even at the low end of this prevention to cost ratio, Federal Government experience indicates that automated monitoring should be in place in every department and agency, and particularly those like the Centers for Medicare and Medicaid Services (CMS), the Internal Revenue Service (IRS) Earned Income Tax Credit (EITC), and Department of Labor Unemployment Insurance (UI) programs that are at the top of the GAO’s “high priority programs” list.  Further, achieving President Obama’s objective of attacking fraud, waste, and abuse in government spending means applying automated monitoring to all payments including vendor payments, grants, purchase cards, fleet cards, and travel cards.

Most people now agree that “pay and chase” activities like recovery audits aren’t the answer.  In both tough times and prosperous times it seems like common sense to ensure that improper payments are prevented before they occur.  And when the worst case return on investment is 3:1, automated monitoring seems like a solution with which both sides of the aisle can agree.


 

2e91

Business Action – Powerful Concept When Delivered to Frontlines

On the subject of industry analyst Dr. Robin Bloor’s blog “Operational Intelligence – Is This the End Game for Big Data?”, I would offer that “Business Action” is a powerful concept when delivered to the frontline employee. The strategic decision making process of management and executives has been well served by traditional BI; however, the needs of the frontline employees require a different model. In today’s business environment the frontline employee doesn’t have time for ad hoc analysis and they may not have the analytical skills (or access to a consultant or analyst as an executive does).  However if they are given an insight at the right time and in the right form then they will readily make the smarter decision.

In our experiences there are several key attributes for a “business action insight” (BAI).  Timely in the goldilocks sense is important – not too soon and not too late but instead at the time when the decision needs to be made. The BAI should be pointed – specifically relevant to the task and decision being worked on. It must be easily absorbed – clear as to why the analysis produced the insight and with supporting data immediately available. In regulatory domains traceability of the insight is necessary.

We’ve seen tremendous paybacks from driving smarter decisions at the frontline.


 

The Devil is in the Details

devils-in-the-details-graphicAs it regards to the recent Smart Data Collective piece, “Big Data Analytics, Business Intelligence and the Mind of Sherlock Holmes”, whether we’re talking Sherlock Holmes or the premise of the popular television show, The Mentalist, the key is in the details. The “Big Data” revolution is all about keeping the detailed information so that it is available for examination.  And that’s driving a revolution in BI.  Before we looked for problems in a KPI, if the KPI was off enough we dig in and try to find the cause of the problem.  There’s a new “inverted BI” we’re seeing at Oversight. Think of the classic BI process of noticing a KPI is off and drilling and asking questions to uncover the factors that are influencing the KPI.  You’re engaging in a process to discover the indicators of the core problem.  And you drill down into ever more detail until you’ve found the problem.  Inverted BI turns that process upside down.  With Big Data you can continuously comb through the details looking for the indicators of the issue.  The real advantage is things don’t get lost in the average of a KPI.  By leveraging the detailed information available with Big Data we can identify the issues at the detail level.  We’re turning the whole drill down notion on its head.


 

2920

What the Obama Campaign Teaches Us about Big Data Analytics

What the Obama Campaign Teaches Us about Big Data AnalyticsThe election is long over but I find it interesting to look at some of the ways that the Obama Campaign successfully leveraged data. The real trick to a Big Data Application (BDA) was demonstrated by the Obama campaign in the way they closed the last mile of the election.  Yes they united many different sources of data and yes they analyzed it in many innovative dimensions. But the thing that really drove the value was the fact that the campaign staffers knew exactly who to visit and what to say. When they went to visit a voter, they had a briefing sheet that told them what to talk about to that person, what points were likely to resonate, what would motivate that person to get out and vote.  They knew which voters to call on, in terms of voters that they could swing to the President and they knew how to do it.

To put in a business context, it’s more than the sales reps knowing all the features and benefits of the product lines. We can use Big Data to help them know the prospects most likely to purchase a product AND the benefits that would mean the most to that specific prospect.

There’s really too much focus on the mind blowing “I never would have thought of that” analytics.  The thing the Obama campaign taught us about business analytics is it’s more how you put the data and the analysis to work that matters.

In my opinion, that’s the flaw in so much of the Big Data dialogue, there’s too much emphasis on the “gee whiz analysis” and “look at this really cool graphic”.  Instead it should be on how we put a big data insight to work to collectively move the needle.

The real value in big data is the fact that we have all the details, the peculiar and subtle insights are where the unique advantages are. It’s the fact that we don’t’ need to deal with averages and the inherent loss of fidelity created by the averaging process that creates the compelling advantage.

That’s what we need to focus on – what can we find in the details that can let a front line employee make a smarter decision – just like the Obama campaign workers.

http://www.datanami.com/datanami/2013-01-04/big_data_applications_not_meeting_expectations.html

Beyond BI: What’s in a Graph – Really?

Beyond BI: What’s in a Graph - Really?

While perusing a BI demo I was struck by a featured graph with 15 bars showing sales results for various promotions.  The presentation was beautiful and the tooling impressive.  Then I started wondering what information I should glean from the chart.  The bars varied around a mean, some higher, some lower.  Were the laggards a significant problem?  Were the winners a sign of future success?  Would I have ever noticed these outliers in the sea of dashboards presented to me every day?  With enough time and effort, perhaps the answers to these questions is “yes.”  But the very success of BI systems redoubles the need for a tool like Oversight, which does such analysis for me.  Continuous Transaction Analysis handles the manual labor of reviewing reports and presents only the most important insights to me.  Of course, I can review the findings and dig deeper if need be.  But given the constant onslaught of information and increasing demands on my time, any help with picking the wheat from the chaff is greatly appreciated.

2221

The Messy World of Big Data

Gartner analyst Svetlana Sicular recently wrote on the topic “Big Data is Falling into the Trough of Disillusionment.” It provided much food for big data thoughts. And, in a world of uncertainty we have to think in terms of many questions with the answer from each question potentially identifying an indicator of an issue. When we’re outside the highly structured word of traditional BI there’s no longer the one question that provides the black and white answer. To build useful applications in the messy world of Big Data we need to think of the range of indicators that “perhaps” can identify an issue.

In many ways this is similar to the challenge of identifying fraudulent business transactions. The fraudster attempts to conceal the crime which creates the messy analysis situation.  Fraud was our original design point for Oversight’s platform.  From those lessons we’ve developed a general purpose platform that contemplates a multifactor analysis, one that searches detailed data for multiple indicators.  This enables the type of application that will get us to the plateau of productivity.

Real-Time Analytics and the Importance of Timeliness

SearchBusinessAnalytics recently published a very timely piece entitled, “Real-time analytics brings BI data directly into business operations”. The article articulates how businesses can take advantage of analytics tools to gain strategic insights and drive real-time operational decision making. What we have been talking about as continuous analysis fits squarely into this camp; delivering these benefits. An evolution of business intelligence, real-time analytics can also be thought of as operational intelligence. However, this intelligence is meaningless unless it can be acted upon and acted upon at the right time.

The important notion here is timeliness – you want the insight from the analysis when you’re in the position to act, at the point when you’re ready to make a decision.  The closer you get to driving insights to the frontline the more they need to be immediately digestible, it doesn’t need to be something that has to be interpreted, it needs to be in plain language, easy to understand. The logic behind the conclusion should be readily apparent so the human can add any value from the data they may know that the analytics didn’t. Read on for the full story:http://searchbusinessanalytics.techtarget.com/feature/Real-time-analytics-brings-BI-data-directly-into-business-operations

2ee8

IDC and SAP Whitepaper Offers Valuable Big Data Solutions Overview

The IDC /SAP whitepaper, “Big Data: Trends, Strategies, and SAP Technology. presents a solid case around the emerging shift from the information economy to “the intelligent economy,” where “it is not only access to data but the ability to analyze and act upon it that creates competitive advantage” for businesses. Undoubtedly big data can present overwhelming challenges and there are many popular myths associated with big data (we’ll save that topic for a follow on post) that need to be debunked. However, it is evident the true value big data can reap presents unlimited opportunities for business success. Ultimately, we at Oversight are centering our efforts on the direct knowledge that big data solutions dramatically improve decision making and provide greater insights, faster, to decision makers at all levels of the organization. Delivering insights is at the core of what we do at Oversight and we are delivering solutions today that are enabling companies to utilize big data insights as a means of success.

The white paper also describes IDC’s new Decision Management Framework as it breaks out big data technologies according to their most effective roles and use cases. We think this is a great way for companies to think about and evaluate big data technology opportunities across the spectrum of various decision types and decision makers. The spectrum ranges from strategic decisions to tactical decisions made by either frontline employees or systems with operational decisions lying in between these two endpoints of the IDC decision management framework. As businesses move through the spectrum from strategic to tactical decisions, the greater their need is for automating big data analysis and getting the right insights to the right person at the right time.  The framework also considers the different data and technology needed to collect, monitor, manage, analyze and disseminate actionable insights in order to support the various decision types and decision makers.

According to an IDC survey conducted in the beginning of 2012, the most frequently mentioned Big Data challenges were identified as “deciding what data is relevant” as well as the “lack of appropriate analytics and IT personnel as challenges that inhibit their ability to take full advantage of opportunities presented by Big Data”. As part of organizations’ overall big data strategy, the Oversight platform automates the process of gathering data from lots of events, transactions and behaviors, as well as understand what is relevant and delivers immediate, actionable insights to the frontlines. This puts people to work for positive change within the organization instead of having them spend time they don’t have trying to gather and analyze complex amounts of data.

Cangemi headshotA new video featuring expert Michael Cangemi, former president and CEO of Financial Executives International (FEI) and finance director of Hartstone Group, talks about how continuous monitoring is a critical component to companies’ big data analytics strategy. The main take away is that by continuously monitoring, analyzing and taking action on recommended insights you can better combat many of the threats to your business (bribery, fraud, corruption) as well as maximize profits by making better and faster business decisions.

3 Big Ideas

#1. Continuous monitoring is a TRANSFORMATIONAL TECHNOLOGY.  It goes beyond its more narrow roots of just being thought of as a control application or compliance tool to an enterprise platform that can be integrated into the core of your business. Applied as a comprehensive big data predictive analytics platform across the enterprise, it can be used in a variety of ways to extract actionable insights out of big data that help move the needle in business.

#2. Continuous monitoring impacts GROSS PROFITS. Companies are applying continuous monitoring and big data predictive analytics in creative ways that impact the bottom line. Michael gives us a detailed example in the video, explaining how companies can specifically tighten their process of identifying and understanding deviations in their discounts to improve gross margins.

#3. Continuous monitoring not only helps companies achieve effective FCPA compliance; it is CRITICAL TO MITIGATING FCPA VIOLATIONS. Companies can look at their data on a systematic basis and provide evidence of best practices, which are then rewarded by the SEC and DOJ based on recent cases such as the Peterson case involving Morgan Stanley.

294b

asus

0