Considerations in Forecasting

In my role as forecast analyst,
a day doesn’t go by where someone doesn’t ask a question like the
following, “did you account for X in your Y forecast?”  The answer is
almost always in the affirmative as our methodology incorporates a lot of
research and pulls in a wide range of perspectives.  However, producing
reliable and actionable forecasts is an exercise in prioritization so not every
influencer becomes explicit in the model. 


In our recently completed Portable
Devices forecast
, for example, we evaluated new product features such as touch
screens and the current state of the economy among other considerations. 
For touchscreens we concluded that it encourages product sales but not
necessarily product adoption. So,  people
may upgrade or replace a portable media player with a touch screen device but
someone who has no interest in a portable media device is unlikely to buy one
because of the touch screen interface.  As a result, we thought about the
impact of touch screens and other product enhancements as we projected future
sales but did not include any explicit lever in setting the numbers which
relates to that.  By comparison, in thinking about the economy we
concluded that the general rate of replacement or product ownership lifetime
was likely to change so we modified a calculation to produce a resulting
adjustment in the numbers.


In our forecasts most
influential issues are evident in the deliverable, so there’s visibility to
unit sales, the installed base of devices and current users in this case, but
to really understand all the elements and insights that went into a forecast you
need to talk with the forecast analyst. Very likely, they will answer your
questions “did you account for X in your Y forecast?” with a wholehearted

EDIT - link updated 10-22-09