Paul O’Rorke

Tom Fawcett, Machine Learning Architect at Proofpoint, gave the San Francisco Bay Area ACM Data Mining SIG an insider’s view of email filtering on Monday, October 25th, 2010.  Proofpoint has thousands of customers large and small and guarantees in their service level agreement that customers will get no more than one spam message per 350,000 emails.  Tom pointed out that research on spam filtering has little to do with what companies do in practice in the “real world” and then he revealed a lot about how commercial spam filtering works.

(continue reading…)

Peter Farago and Sean Byrnes gave a juicy and surprising presentation about Flurry’s mobile app analytics at the SDForum Business Intelligence Special Interest Group meeting on 10/19/2010 in Palo Alto.  The title of their presentation was:  ”Your Company’s Mobile App Blind Spot” and it provided both business and technical insights.

Flurry made a big splash in the news when Steve Jobs got pissed off at them and called them out by name in an interview because they outed Apple’s iPad when it was still a closely guarded secret. (See a short video outtake of the interview at VentureBeat.)  Apple responded by changing legal agreements to exclude some third party analytics and some advertising.

(continue reading…)

Salesforce’s CRM analytics architect, Donovan Schneider, presented an overview at the SDForum BI SIG meeting on May 18th, 2010.  Salesforce’s view of analytics is that it should deliver insight that is accessible to mere mortals, real-time, and trustworthy.

(continue reading…)

The book “Who: The A Method for Hiring” describes how companies can stop using voodoo hiring methods and start using more rational and systematic methods. It is based on thousands of interviews of CEOs and other hiring managers and is based on the idea that you should hire specialists, not generalists: the right person for the right spot at the right time.

When hiring, consider what do you want the person you hire to accomplish? Create a scorecard comprised of mission, outcomes, and competencies:

• mission: the job’s core purpose; why the role exists (in 1-5 sentences)
• outcomes: what must get done for A performance (not what the person will be doing but rather 3-8 specific objectives)
• competencies: arbitrarily many role based competencies that describe behaviors that must be demonstrated to achieve outcomes plus 5-8 that describe the company culture

The scorecard ties the two legs of the “A Method” for hiring, “Source” and “Select” together. (continue reading…)

James Taylor, CEO of Decision Management Solutions, gave a talk on “Performance Management and Agility” at the monthly meeting of the SDForum BI SIG on Tuesday, April 20th. He argued that traditional BI and performance management result in dashboards that measure and monitor like instrument clusters in cars. But what is needed is something more like the cockpits in airplanes: there should be buttons and levers and so on that enable the “pilot” to act on the information presented by the dashboard. James argued for combining performance management with decision management (a term he pioneered) so that information supports decision-making that leads to action.
(continue reading…)

The first SDForum conference on analytics, “The Analytics Revolution,” was held in Mountain View on Friday, April 9th, 2010. The conference focused on recent advances in analytics, new opportunities afforded by these advances, and ways companies can take advantage of the analytics revolution in progress.

(continue reading…)

The “Competing on Analytics” panel at the SDForum Conference on “The Analytics Revolution” included people from companies using analytics to “compete at the highest level” according to the five stage maturity model in the book “Competing on Analytics: The New Science of Winning.”  The panelists (Amr Awadallah, Cloudera; Joshua Klahr, Yahoo!; James Phillips, Northscale; Joydeep Sen Sarma, Facebook) represented a good mix from the relatively new Twitter to the larger, older, more established eBay.  David Steier, PriceWaterhouseCoopers, moderated the panel.

(continue reading…)

The panel on “Analyzing Big Data” at the SDForum Analytics Conference on “The Analytics Revolution” included representatives of two companies that analyze data on a petabyte scale (Joydeep Sen Sarma, Facebook and Joshua Klahr, Yahoo!) and two software companies that stand behind open source infrastructure components that are often used to build analytics platforms (Amr Awadalla, Cloudera/Hadoop and James Phillips, Northscale/Memcached and Membase).  The moderator, Owen Thomas of VentureBeat, started off by asking the panelists whether “big data” is a Silicon Valley phenomenon that will soon spread to the Fortune 500 and the rest of the world.

(continue reading…)

Ronny Kohavi (Microsoft) started out by telling a famous true story about Greg Linden’s experience moving a recommender to the shopping cart at Amazon.  A Senior VP of Marketing vetoed Greg’s proposal fearing that it would distract customers from checking out and paying for the items already in their shopping basket reducing conversion.  This is where the “HiPPO” in the title of Ronny’s presentation comes from.  It stands for the “Highest Paid Person’s Opinion” and sometimes for the person (e.g., the VP) holding the opinion.  The Amazon story had a happy ending because Jeff Bezos had established a corporate culture that allowed for experiments to be run so Greg was able to run an experiment to test the hypothesis of the HiPPO.  It turned out that conversions did indeed drop but the increased revenue due to customers purchasing recommended items was substantially greater than the loss.

(continue reading…)

Peter Norvig focused on a major lesson learned at Google and elsewhere in recent years and gave a fascinating keynote presentation on “The Unreasonable Effectiveness of Data” at the SDForum conference on “The Analytics Revolution” April 9th, 2010. The lesson is that data can be surprisingly effective: it can be used to get better performance improvements than one can get from improvements in algorithms.
(continue reading…)