This week, Barack Obama uses analytics to help win the 2012 U.S. Presidential election, and an IBM predictive analytics commercial helps the New Orleans Saints defeat the Philadelphia Eagles on Monday Night Football! As usual, here are my favorite business analytics stories from the past week, and why I think they’re important.
CNN reports that analytics played a HUGE part in Barrack Obama’s victory over Mitt Romney.
Using data in political campaigns isn’t new. BUT, as I understand it, in a lot of races, the methods aren’t much more advanced than slicing and dicing callsheets based on demographic data. Not here. According to the article, an analytics professional (and former Accenture analytics expert) named Rayid Ghani helped make it happen. Mr. Ghani, you’re about to be an analytics rock star.
Now, we don’t know for sure that Mr. Ghani was the chief of the team – but as the only person named in the article, I expect that he was. And we don’t know what tools the campaign used, but Mr. Ghani’s public resume lists “Perl, C, C++, Java, SPSS, SAS, KXEN, and Weka” – all of the tools I’d expect to see listed. He holds an M.S. in Knowledge Discovery and Data Mining from Carnegie Mellon. And surely this has been going on in big races for years – but I haven’t seen alot of detail in print.
Ten bucks says we see a movie about this team two years from now.
Okay not really. But I LOVE IBM’s analytics commercials – and they ran one during this week’s Monday Night Football game.
In this one, IBM describes a very specific problem – an undisclosed “car company” notices that brake lights on their cars are failing more frequently than they should. With a little causality analysis, they figure out that the problem is most common in cars with sunroofs. As it turns out, a problem with the sunroof was damaging the cable connecting the control system to the brake lights and “they didn’t have a brake light problem, they had a sunroof problem.”
In my last post, I complained that reporters and vendors writing about analytics weren’t nearly specific enough. Well, I’m sorry IBM – this is exactly what I think our industry needs. Engaging materials that help potential cliens understand *exactly* what we do.
(And for the record, the Saints did win. Queue the lecture about not confusing correlation with causation.)
The Forrester WaveTM: Business Intelligence Services Providers, Q4 2012 report is out, and IBM does well.
Yep, another press release. But this one has a link to the Forrester BI report, including great statistics on the number of business intelligence projects (and at what revenue) are being run by IBM, Accenture, Capgemini, etc. I agree – IBM is doing well. Business Intelligence services, Predictive Analytics services, and high performance computing hardware are all required for a good analytics project; IBM can do all three. Plus, I really like their commercials.
Apparently, Tableau 8 is in beta – and has some interesting new features, according to ZDNet.
It’s no secret that I like Tableau software for situations where I need to create good looking graphics and dashboards RAPIDLY. At their recent user conference, they demoed a beta of Tableau 8. The good news? Treemaps! In-browser editing! And connectivity to Google Analytics and Salesforce.com. Oddly enough, those were on my wish list. Now, if only I could perform an ANOVA or a Decision Tree….