The degree of exactness of places on a GIS map depends on what geographic level you are looking for, such as county or an address. A specific place needs a “clean” address: street, city, state, and zip code. “Clean” means a database address with no problems, errors, or limitations which can result in places displayed in wrong locations, or not displayed at all.
For clean address data and minimized mapping errors:
- Use the true, complete name of the street, such as “2140 Front Street,” not just “2140 Front”
- Include “directions,” such as “123 East Main Street”
- Use standardized address information, such as correct abbreviations, “St,” for “Street”
- Include the street number, not just the street name
- Spell out street addresses and cities, such as “Martin Luther King” instead of “MLK”5
- Do not use “Post Office Box” numbers
- Do not include extensions in database fields for street addresses, such as an “apartment number,” “suite,” and “building name” (which should be in a separate database field)
- Make database address fields “text” columns, not numeric
- Make sure database fields are complete (no missing data)
- Use a unique “ID” field for each database record, with no blanks and no duplicated IDs
- Update database addresses periodically
- Use GIS geocoding software that has the latest streets (such as new subdivisions) and the best possible reference data sets
- Do a quality control on your final geocoded data by examining a sample.
Do not use zip codes for geocoding
Zip codes for geocoding have limitations and are only approximations. Zip codes are mail-delivery routes, not areas with exact spatial boundaries. Zip code assignments are linked to mail volume, delivery area size, geographic location, and topography, but not necessarily to community boundaries. Some zip codes do not have contiguous areas, so the same zip code may represent addresses several miles apart, with other zip codes in-between.
Completeness and accuracy
Completeness is the proportion of addresses that are successfully geocoded. If many records in a database can’t be geocoded, mapped data may be misleading. Different reference data sets vary in the geocoding software algorithms used to match an address to a street segment.
Although geocoded data may have completeness, accuracy may be more important. For example, a database may have an address at 100 Main Street, Middleburg Heights, OH, and 100 Main Street, Middleburg, OH. A geocoding system might consider both to be the same. Although both addresses were geocoded, one of them is inaccurate.
If your business maps are incomplete or misleading, or if your database is not yielding productive results (such as too many undeliverable target emails and postal mails), let me tell you more about clean addresses. And then create a results-oriented map for cost-effective mailings, better sales territory assignments, and more productive sales calls.