Forecasted Sales

1. A few recommendations would serve very useful to Mr. Perez for developing more precise sales forecasts. First, the same method should not be used for each of the chemical group’s division. Because the market is so different in each division, some techniques will not work for one division like it would for another. For example, a test market may work for a division with a brand new product, but it would not be as helpful or efficient for a division where there has been steady growth and established products for the past five years. The other suggestion I have would be to not begin the market analysis with reviewing the sales in each division of the company. If he would like to know more about the market, it would make more sense to use the build-up approach to conduct primary research. It would benefit the company because certain divisions are dependent on the oil market which ties into my last suggestion that he should gather statistics and forecast the standing of the crude oil market.

2. Each division requires its own specific type of forecasting method because of all of their unique circumstances. For the industrial intermediaries division, polymers and detergents each need a different method. Since the polymer intermediates sales have risen by a high number of 40% over a two year period and are unpredictable, an exponential smoothing method would be most appropriate. This method weights highest on the most recent years sales history. For the detergent intermediates, a moving average method would be most appropriate because sales have been stable overall in the past years. There is no need to give the most recent periods a higher weight of importance. For the specialty chemicals division, it would be most effective to complete an econometric model. Determining how drastically a change in the oil market would affect the sales in this division would be very helpful for the overall sales forecast. Finally, the bromine division should be forecasted using the test market method. Since there is no past history to go by, a test market would allow a full dress rehearsal in a small market area. It is also a good option because there are very few competitors. If worry about competitors disrupting the test market occur, AKAMAI could even move the test into a private laboratory to ensure accuracy of the test.

3. With the data given, it would not be possible to complete an econometric model for the specialty chemicals division because there is no information about the crude oil conditions. Although, because the corporation has a very good market position in this division, I believe that an exponential smoothing forecast will still be suitable. My forecast is as follows: 1/9(50 sales’02) +1/8(60.6 sales’03) +1/7(89.1 sales’04) +1/4(124.6 sales’05) +1/3(153.5 sales’06) + 1/2(191.9 sales’07) = 233.1 Forecasted Sales for 2008 I decided to place the least weight on the sales from the later three years and more weight on the past three years because the market grew the most in these past three years.

4. When forecasting the industrial intermediaries division, it would have been helpful to have the separate sales summaries for the polymer and detergent. The intermediates have very different market conditions which means forecasting them will require different steps. Only the total industrial intermediates sales summary is available. Having separate sales summaries would allow for more accurate forecasting to take place. I would use a moving average method for detergent and an exponential smoothing method for polymer.