OK, I admit it—I have been a good boy for most of my life. I did my homework, didn’t get into trouble, and used the recommended corporate IT applications. But I am wondering if I—and all the analytical people I work with—need to break out a little. Maybe we need to become analytical bad boys and bad girls.
I am prompted to address this thought by a couple of things I’ve read lately. One is the mildly revolutionary blogging of Chuck Hollis at EMC. Chuck is leading a social transformation of EMC, and his recent blog on analytics (http://chucksblog.emc.com/chucks_blog/2010/06/the-coming-revolution-in-b...) casts doubt on several aspects of traditionally responsible analytical practice. Should you wait for the data you want to appear in the enterprise data warehouse, or hack something together? For that matter, should you plan and build a big EDW, or just pull together a series of data marts? Should you spend years toiling away on Master Data Management, or just get something done? Hollis makes a convincing case for the latter alternatives. If an executive at a once-conservative company making enterprise storage devices can be radical, you and I can too. For God’s sake, I’ve even got tenure!
Beyond the analytical sphere, I’ve been reading a subversive manuscript called Hacking Work by Bill Jensen and Josh Klein (conventional practice would suggest that I not mention the book until it’s published, and it’s not coming out until September—but they encourage subversion, right?). The book correctly points out that if you want to get anything done, you often have to work around corporate practices and procedures. They have an excellent chapter on hacking corporate IT practices. Right now I am happy to be using an unauthorized Babson computer. When the book comes out, I’ll have more to say about their take on analytics—and to warn you or whet your appetite, it isn’t entirely complimentary.
My conclusion from these readings is that we need to be a little more insurgent ourselves in getting our analytical work done. It’s pretty easy right now to compile the data you need, buy and use the software you need, and disseminate the results you need to spread—all at prices that can be disguised in most corporate expense policies. What has been labeled “the consumer IT revolution” needs to spread to analytics. If you have good intentions and get your work done efficiently and effectively, things will probably work out OK in the end.