Budgets, Models, Simulations and Whirlybirds
As we get into budget season and budget models fly around wildly, and almost always out of control, a client asked me to write a column on the subject of models and simulations. Since there are too many root causes for a bad model outcome to cover in a single entry, I thought I would start a series to share some common modeling traps operators should be aware of before accepting the output of any model. That applies to accepting the budget model from finance, signing on to operating a new organization born through an efficiency model, launching a new product with a revenue forecast based on a marketing model, or taking on the P&L responsibility of a new merged or acquired entity that looks good on paper and in the models of the investment bankers.
It has been my experience and that of my partners, that many mergers, acquisitions, large outsourcing deals, and product launches failed to produce the desired or forecasted results because the teams working on those transactions had a modeling focus that was not grounded in operational reality.
Unfortunatly, that experience has caused some of our operator clients to discount the output of models and the value of simulations and focus more on gut feelings, instincts, and traditional budgeting analysis. That’s a shame because models and simulations are extremely important tools that can add significant value in operating a business. Properly designed, they help validate, temper, and fine-tune our decisions by identifying gaps in our thinking and highlighting hidden but critical assumptions we are basing our plans on. Models help analyze alternative scenarios quickly and can highlight potential risk points along the path of a plan through simulation. Modeling and simulating the lifecycle of an M&A transaction, an outsourcing contract lifecycle, a new product’s market launch, or a new operating strategy upfront can save tremendous amounts of energy in the back end. Alternative scenarios can be analyzed in a few minutes or hours on a computer rather than through trial and error over a few years in the market place, and assumptions can be altered in real time to assess their impact over the lifetime of a project.
In my experience over the years, I have found that if I can model or simulate anything before I actually do it, the investment always pays off large dividends. In our consulting practice, as well as when I run operations as a contract executive, models and simulations are important tools in my toolbag. Taking a hint from the military and commercial airline pilots, I even use a flight simulator for my remote control helicopter hobby (or whirlybird addiction as my wife calls it). Building an RC Heli can take dozens of hours and cost from a few hundred to a couple of thousand dollars. They look cool and pretty when I finish them, but learning the unique aspects of flying each one of them is an exercise that involves frequent crashes until flying them becomes second nature. Learning to fly one of them on the computer simulator is a heck of a lot cheaper - and takes less time to be airborne again - than having to repair one after a crash, let alone having to endure my wife’s ribbing about my piloting skills. Similarly, I prefer watching the top line of a business take a nosedive because of a bad assumption, or an outsourcing transaction not generating the expected savings because of a flawed process map in a spreadsheet or forecasting model rather than real life.
The process of creating baseline economic and operating models is really not that difficult and with a little practice most of us can do it. For example, rather than using any of the really complex modeling tools, I generally create a few basic spreadsheets in Excel and then build real life dashboards from them using the simple modeling tools that come with Excel or add-on tools such as Business Object’s Xcelcius platform. It does take a little forethought to construct a spreadsheet that can be used as input to a model, but it certainly is not brain surgery or rocket science.
Some people prefer more complicated tools, but I found they require extensive training on the tool itself, intimate knowledge of modeling, and are generally beyond the skills of mere mortals, business executives, operators, and other folks like myself. Unless they require highly complex multi-dimensional forecasting, in which case they should really have experts do it, or have a passion for it, in which case they are probably in the wrong job, most operators derive little incremental value from training on how to use those tools.
Of course modeling tools, whether the simpler ones I choose to use, or the models used by sophisticated market forecasting firms and financial departments of global corporations, are only as good as the inputs that go into them. Over the last fifteen years I have worked on quite a few projects and companies that needed intervention because the actual results did not match the modeling output that drove the original decision. Most produced unexpected and, most of the times, disappointing results that required an operational or technology turn-around, or at least a merciful execution.
The focus of these series of articles is to highlight some of the common mistakes modelers make when they are not connected to an operator.