Guest post from Kurt Peckman, Program Director, IBM Business Analytics
Follow Kurt on Twitter: @KurtPeckman
AP** Wire – Shareholders are urging Santa Claus, CEO, to take a temporary leave of absence for publically divulging the details of the North Pole’s newly acquired big data platform to Graeme Noseworthy, a marketing director at IBM.
Codenamed Operation Kringle, the big data platform enables the North Pole to execute advanced analytics “…for profiling, manufacturing and distribution in less than 10 seconds” versus what used to take hundreds of elves more than three weeks to complete with rudimentary spreadsheets.
Operation Kringle provides true competitive advantage for North Pole operations, and shareholders demonstrated their anger over the public relations gaffe by a sell off of North Pole [NPOL] shares.
News of the intellectual property breach and sell off still ripples thru the “Stocking” Market. NPOL closed down 19 percent in after-hours trading yesterday, which is newsworthy given that only about one minute around the world is considered “after hours” relative to the North Pole.
Mr. Claus revealed so much insider information about the big data project that SEC investigators have dubbed the set of responses in the blog post “The Letter.” The Letter outlines detailed descriptions of Operation Kringle, including: Offline Sibling Sentiment Data, Elven Internal Call Center specifications, and real time Naughty/Nice Child Data (NNCD).
The fallout follows on the heels of an initial gaffe during a Dec. 6 IBM interview with “Fiscal Cliff” Clavin. In that interview Mr. Clavin explained the components of a quality decision management system: constraint-based optimization, business rules and predictive analytics. Mr. Clavin pointed out that a primary goal of such systems is to operationalize predictive intelligence by automating high volume decision-making out on the front lines.
During that interview, Mr. Clavin let slip that he was working on a cloud-based analytics solution that organizations around the world, including the North Pole, are leveraging to “dramatically improve operations” through better decisions and answers, at a lower cost.
The Lead Elf on the cloud analytics project, Hermie, declined requests for an interview although did forward this statement though North Pole attorneys, and also deferred any other questions to IBM’s Colin Shearer (watch Shearer discuss analytics in the cloud):
“Naughty/Nice (N/N) conversions were a real problem area which is why we decided to explore a cloud analytics solution. Santa, Jack [Frost, VP of Operations] and I all embraced a recommendation to move (N/N) predictive scoring into the cloud.
Before implementing IBM Analytic Answers, we were leveraging the NNCD database and locally developing models to segment naughty and nice children. We used several types of decision trees to drill down into the attributes of naughty versus nice, and text analytics to build predictive N/N comment models.
Did you know that a child who asks, ‘How was work today?,’ is 2.7x more likely to be naughty within a 6.1 hour window? By the way, it's amazing what correlations predictive analytics can uncover.
Finally, we used all results to predict N/N categorization with a very high degree of accuracy.
Our number one objective for using the Cloud is to answer a single question, ‘What is the likelihood of a conversion from N/N?’ Managers easily upload a .csv list of naughty children and are then able to quickly download the list of those most likely to convert to nice before the sleds go out.
By moving to IBM Analytic Answers, we anticipate the coal return rate from Nice children to drop 40 percent for a target savings of $900,000 (or $15 million Keebler cookies).
Overall, the elves deeply appreciate everything Santa has accomplished with Operation Kringle. Things have gotten a lot better, morale is up, and we are much more efficient.
It’s getting harder to stay out in front of children given their ability with technology, so we will be relying more and more on Operation Kringle and IBM Analytic Answers to move us forward.”
** “Analytics that are Predictive” News Services