I'm back. Hope things have gone well for all of you. I made significant progress on all of my exams (still ongoing) and class projects (also, still ongoing). Anyway, I promised you a new entry in my forecasting series, so today we are going to look at Trend Projections. With it, you can see if your sales are likely to continue upward, or if they are falling over a cliff.
When you have a series of past data, you can plot them on a graph. Then, using the least squares method, you can find a line that will best fit those observations. Draw that line out into the future. Technically, the least squares methods finds the line in which there is the smallest sum of squares in the absolute difference between all of the data points to a particular linear model. If you want more information about it, you can read about it over at this MathWorld link HERE.
The linear model follows that same linear model you probably learned back in Junior High School:
y=bx + a
with a=the y-intercept
x=independent variable of time.
y is the value that we want to predict.
So after you draw out your line, you can predict what the most likely plot will be along that line. Now, of course real life and real sales doesn't always occur in a linear fashion. To help deal with that, you could figure out what the standard error of the estimate is. Simply put, that's just creating a bell curve about the next projected plotted point. This will tell us the likelihood that our next period demand or sales will fall in a particular range of values, and forms a sort of band about the linear trend projection.
Hopefully, I've given you a good starting look into the use of trend projection. I intend to look more into linear regression and standard errors one more time in tomorrow's blog post. Statistics isn't my strongest area, but I will try to explain it to the best of my ability. See you then.
--------Sincerely, Trevor Stasik.
To return to initial post about forecasting, click HERE.
To visit the next post in the forecasting series, click HERE.
Trends, Least Squares Method, Projection, Forecasting