Structure in Threes: Capacity Planning
September 2, 2013 Leave a comment
Approximately two years ago I was just finishing an engagement designing and building a IT Shared Services Service. It was becoming the prototype for other services that would be offered in the Shares Services organization. Some of the difficulties with designing a new service for the first time is that most people in the organization are still thinking they are to be organized and execute in the same old model. So when I stated discussing a complete serviced; it became an RCA Victor dog look moment; they were hearing words but clueless on the meaning. Once I had a breakthrough on how the service function would be organized the next problem to tackle was capacity planning.
Typically capacity planning had been a buy y capacity with x% in reserve during the initial year’s budget and hold the line against user requests for additional capacity. This works if you have an IT fixed budget, that is there will be no additional funds allocated despite growth in demand. However, how often does that happen. If you’re going to have a services organization demand is likely to grow and recede with work requirements. This suggests the old once a year planning cycle is not agile enough for an enterprises needs, nor the old methodologies.
The solution I suggested involved previous experience in marketing management; where the ancient voodoo art of demand forecasting and management is practiced. I took the forecasting techniques marketing uses as well as the historical capacity monitoring that IT uses and blended them together into a system that tracked and forecasted capacity demands based upon current utilization trends and requests for service. This forecast was send to the operations portion of the service to give them early warning as to when they would likely cross a threshold of capacity where they needed to order more or restrict utilization. The order management component [request for new service would created demand for x% more capacity] of this system was implemented with fairly good results. The other portions were left to be implement in another project, due to budget cutbacks and the necessity to create and internal Business Intelligence function. However, I left at the end of that year for other opportunities closer to home, but still have the logic for doing this on my white board in my home office.
I thought it interesting when I discussed the approach with some peers at work; however, when I mentioned CAGR for capacity planning, there was that RCA Victor dog look moment again. I’ve since run multiple industrial dynamics simulations and the logic proves out. I’ll be looking for a forward thinking CIO in the next couple of months to beta test the methodology as part of my due diligence for the tools and techniques I put in the book. So far all the methods, templates, and strategies I’ve marked for inclusion have been field tested for practicality over the past 20+ years.