Geocoding


Geocoding Customer Addresses

Problem

A local business in Raleigh, North Carolina wants to develop an advertising campaign targeting potential customers in underrepresented areas.  The business has current customer data in Excel.

Analysis Procedures

The overall strategy was to use the customer data to geocode first ZIP Codes and then street-level addresses using the geocoding tools in ArcMap.  The strategy involved creating several new layers.

First, I imported the customer data sheet into ArcMap.  After reviewing the data table, I created an Address Locator using ArcCatalog.  I then linked the Address Locator to reference layer and activated it using Geocoding tools in ArcMap.  The map displayed several points representing customers location at the ZIP Code level.  This method, which is called Batch Geocoding Process resulted in 154 addresses that were matched, 57 were tied, and 44 remain unmatched.  I then manually fixed some unmatched results using the interactive approach.

Next, I geocoded customer addresses using a street layer. The wake County boundary and wake County streets files were provided. The wake County streets file was used as a reference data file. Using ArcCatalog I created Address Locator and followed the same steps I used earlier. After making some parameters modification, I activated the Address Locator and located addresses interactively in ArcMap. I created and used US Address – Dual Ranges to run the batch geocoding process. The process generated a new layer that shows individual customer addresses.

Flowchart



Results

Map 1: Customers addresses geocoded by ZIP Code


Map 2: Customers addresses geocoded by streets


Application & Reflection

Geocoding has many applications in health. One of the applications is using geocoding to analyze the relationship between physician practice sites and socioeconomic factors in Mecklenburg County. The addresses of physician practice sites will be purchased from the North Carolina Medical Society. The information on social demographic factors will be obtained from the American Community Survey ((ACS), U.S. Census Bureau). I will create Address Locator using street-level TIGER files downloaded from the U.S. Census Bureau website.  I will then match the location of physician sites addresses and overly poverty income ratio data from ACS.