In his book, How to Measure Anything, Douglas W. Hubbard laments the reluctance of IT organizations to use inferential statistics for measurement. Hubbard poses the question of how to measure the number of fish in a lake. Many IT decision makers might respond just to drain the lake, kill the fish, then have someone slog through the mud and count the fish. Hubbard suggests another method. Simply go out and catch 1000 fish and tag them. Then release them, come back later and catch another 1000 fish then count the number of tags. Of course, this isn’t as precise, but often the degree of precision is enough to make an informed decision.

 So, what’s this got to with legacy transformation? Remember in my last post I spoke of asymmetrical transformation, transforming legacy code to specialized frameworks and tools such as extract/transform/load (ETL) tools and reporting tools. So how do we know how much ETL code there is in a legacy application? Should we drain the lake again? I suggest taking a page from Hubbard’s book and using inferential statistics. In the example below we randomly selected 259 modules and categorized each module. This took about 8 hours. Better than draining the lake, which would have taken about 230 hours? We have a margin of error, of course, but this is good enough to make some decisions, such as how to proceed with an asymmetrical transformation.

Geoffrey Moore, best-selling author and managing director of leading consulting firm TCG Advisors, calls this differentiation core and context. Core makes your business unique, context is all the useful utility functionality that is required, but doesn’t merit hand-written code. Mr. Moore will be keynoting our conference. Details below. Check out his blog: http://geoffmoore.blogs.com/

 Application Transformation: Getting to the bottom line.  

 Nov 5^th, Americas: http://www.bitpipe.com/hpsolutionsconference/america

Nov 4^th, Europe, Middle East, Africa: http://www.bitpipe.com/hpsolutionsconference/europe

Nov 3^rd, Asia, Pacific and Japan: http://www.bitpipe.com/hpsolutionsconference/asia