REASONS FOR ESTABLISHING SALES TERRITORIES
WHEN NOT TO ESTABLISH SALES TERRITORIES
SOME GUIDELINES FOR DESIGNING TERRITORIES
Following these guidelines will help in ensuring that tours are as short as possible:
More efficient (shorter) routes will tend to exhibit one of these patterns:
Note that the cloverleaf pattern better follows the guidelines that were given above. Indeed, with the example territories and focal point above, the cloverleaf routes would probably require less travel time if such routing is possible on existing roads.
The above circular area was divided into five equally sized territories with a focal point at the center. A route to visit customers in the territories was then drawn in either a circular clover leaf pattern or in a hopscotch pattern. This way of making territories and of routing sales calls would be appropriate if, say, five salespeople reported to a common office in the center. It would also be appropriate if, say, a single salesperson was assigned to a remote territory and must divide the territory into five daily routes to visit customers once per week.
TERRITORIES WITH A CENTRAL FOCAL POINT:
In the above example, all territories met at a central point. This could be a centralized office out of which all salespeople work. This sort of design could also be appropriate for a route salesperson who must visit customers once per week, with a Monday route, a Tuesday route, and so on.
To generate territories of approximately equal workload, divide the total number of customers and prospects that must be visited by the number of salespeople that are available to cover the area. If the area has five salespeople and sixty acounts which must be visited daily, then each salesperson would be assigned twelve accounts if these accounts are evenly distributed across the area.
If there are no natural boundaries, start the process of generating the five territories by laying a strait edge through the central office pointing north. Draw a line directly from the central office to the edge of the circlular boundary. Now sweep the straight edge clockwise, like the second hand on a clock, and count the number of customer accounts that it sweeps across. When it sweeps across twelve accounts, draw another line from the central office to the edge of the circular boundary. Sweep the straight edge through another twelve accounts and draw another line. Keep doing this until five wedge shaped territories have been generated. These should all be approximately equal in workload.
THE TRAVELING SALESPERSON PROBLEM:
It would seem that there should be some sort of mathematical method for developing the shortest tours within a territory, but there is no way to develop the shortest route other than to try every combination or roads possible. If the routes are constantly changing - as when a salesperson has a different set of prospects to visit every day - then this is not feasible. Fortunately, there are some heuristics that can be used to find routes that tend to be reasonably short.
Largest Angle Heuristic
Consider the choice below:
You have traveled from Point A to Point B. Should you travel next to Point C or Point D? Using the Largest Angle Heuristic, you should choose to travel to Point D. When choosing a route using the Largest Angle Heuristic, always travel next to a point which generates the largest angle to the pointed visited last.
This heuristic will help to ensure that the route is circular and tends to generate efficient routes.
Closest Next Heuristic
Consider the choice below:
You have traveled from Point A to Point B. Should you travel next to Point C or Point D? Using the Closest Next Heuristic, you should choose to travel to Point D. When choosing a route using the Closest Next Heuristic, always travel next to a point which is closest to where you are right now.
This heuristic also tends to generate efficient routes. Which is better, the Largest Angle Heuristic or the Closest Next Heuristic? There is no good answer to this question; the only way to find the optimal route to a particular problem is to simulate every possible route on a computer (or by hand) - a very time consuming process. A realistic problem for a salesperson is to generate a reasonably good route for a single day. Using these heuristics, you might not always find the optimal route, but you can quickly rough out a route that generally is relatively good with very little investment in time.
In using these heuristics to make decisions, you should also use the Look Ahead procedure. That is, don't simply look at the next point in a route, but look ahead two points and use try both of the above heuristics to make a decision for the next two points.