The purpose of this lab is to acquaint you with some basic descriptive spatial statistics as one approach to defining primary and secondary markets.
quadrat grid in miles
0 1 2 3 4 5
___________________________________________________
0 | |
| |
| * |
| * |
| * |
1 | * |
| * |
| * |
| * |
^ | |
| 2 | |
N | * |
| | * * * |
| * |
| * |
3 | |
| |
| |
| * * |
| * |
4 | * |
| * |
| * |
| |
| |
5 | * |
--------------------------------------------------
Above is a map of repeat customers to your suburban hair adjustment salon (beauty and barber shop). The addresses alone are shown, without the streets. Measure the X distance from the left border to each customer's home address (you might want to label each of the 20 customers on this map so you can keep track of them). Store your measurements in Column A of a spreadsheet. Measure the Y distance from the top border of the map to each customer's address, and store those measurements in Column B of your spreadsheet. Make sure that each X-Y pair is for the SAME CUSTOMER!
To make these measurements, by the way, measure the X axis and the Y axis on the map below with a ruler (it's probably somewhere around 4" or 10 cm). These axial measurements (which may not be exactly the same because of my artistic limitations in html) represent 5 miles on the ground. So, you measure each location in inches or centimeters from each axis. Then, multiply each measurement by 5 (miles) and divide by 4 inches or 10 cm (or whatever the axial measurements turned out to be) to get the mileage.
At the bottom of columns A and B, calculate the average for each. Your spreadsheet can do this if you type "=average(a1:a20)" and "=average(b1:b20)". Place a contrasting mark on the map at the mean X and mean Y coörodinate. This is the mean center of your customer distribution (which may not coïncide with the location of your store!).
To calculate the standard distance, you will need to create six more columns. It will probably be easiest on your eyes if you insert three columns (B, C, and D) between your X measurements (Column A) and your Y measurements (now Column E). You will also use columns F, G, and H.
In Column B, copy the mean X to B1 and then copy that value down the column. Do the same with the mean Y in Column F.
In Column C, subtract the mean X from the X measurement for each customer. In other words, put this formula, "=a1-b1", in Cell C1. Then, copy the formula down the column. These are the deviations of each customer's location from the mean center's X coördinate. Do the same sort of thing in Column G with the mean Y (i.e., Cell G1 should read "=E1-F1", and the formula should be copied down Column G).
In columns D and H, you're going to square those deviations. That is, Cell D1 should read either "=C1^2" or "=C1*C1" and Cell H1 should read "=G1^2" or "=G1*G1". The purpose of squaring the deviations, by the way, is to get rid of the negative values (so when you average them, you don't wind up with a big, fat, meaningless zero).
Now, sum columns D and H. You now have the sum of the squared deviations.
Divide each of the summed squared deviations by 20 (the number of your loyal grooming customers). That is, in Cell D22, you'd write "=D21/20" and similarly over in H.
Now, add the results of the two divisions. You'll get some big number.
Bring that number back to Earth by taking its square root ("=sqrt..."). This number is your standard distance. Standard distance is the two-dimensional (i.e., spatial) standard deviation.
Arbitrarily, define your primary market area in terms of percentage of your customers included in it (e.g., 60%, 70%, whatever). Divide that percentage by 2, so you can enter the normal curve table (the standard distance corresponds to a 2-tailed pattern). Now, multiply THAT answer by 1000. You're ready for the table (handout). Look for this last answer in the body of the table (get as close as you can). Read to the left axis to get part of the Z score and then up to the top to get the rest of the Z score (e.g., you might find 1.5 on the left and .06 on the top, creating a Z score of 1.56).
Multiply that Z score by the standard distance to get the mileage enclosing your primary market area. Draw a circle on the map, centering it on the mean center for your actual distribution. Its radius will be the mileage defined above, which corresponds to your primary market. Label this circle "primary market." This circle will enclose very roughly your arbitrarily- defined primary market customers (you might want to count the dots in the circle and see if they amount to roughly the percentage you defined as the primary market, just to be sure you're not way off).
Repeat the process for your arbitrarily-defined secondary market (e.g., 70%, 80%, or whatever). Draw and label that circle as the "secondary market."