To illustrate point in polygon processing, consider the map below. Let's say it represents sales territories for an insurance company. It also shows a flood plain that exists in these sales territories.
Now let's say a potential customer contacts the insurance company to request a quote for homeowner's insurance. To generate the quote, the insurance company must determine whether or not the home resides in a flood plain. It also must determine the sales territory where the property resides. To make these determinations, the insurance company must identify the "polygons," or areas, in which the house resides. The house's address is converted to latitude/longitude coordinates. (This could be done using the Spectrum Global Geocode.) These coordinates are then used as input to Point In Polygon. Let's say coordinates indicate that the house is in this location:
Point In Polygon determines that this house resides in the "Sales Territory 4" polygon. Point In Polygon also determines that the house does not reside in the "Flood Plain" polygon, so there is no need to include flood insurance in the quote.
As this illustration shows, you perform multiple analyses on a given location. In this case, you want to know two things: the sales territory and whether or not the location is in a flood plain. To accomplish this, you would need two databases to match against: one database that contains the location of flood plains and one that contains the location of your sales territories. To match against multiple databases, construct a job in Enterprise Designer that has multiple Point In Polygon stages, one for each database.