News - Fresh Data Archive Article

Return to Fresh Data Blog
Return to Fresh Data Archive

Data Analytics and the Pitfalls of the Zip Code

Date: February, 2014 --

The life of an analyst whose data includes the zip/postal code can sometimes get rocky.  A recent story from Level Plains, Alabama underscores the conundrum of data being applied in ways that aren’t consistent with the purposes for which the data was created in the first place.  In short, we need to be sure that the application or use we make of the data is consistent with the purpose for which it was created. Here is a real dollars-and-cents example of this.  

            The USPS Zip code system was implemented in the ‘60’s, but has a longer history.  During World War II, many experienced postal workers were called to serve in the military. These were replaced with workers who were unfamiliar with the structure of the mail sort and delivery system. In order to remedy this, the USPS developed an elementary single digit coding system that was applied to urban areas to enable the new “fill-in” postal workers to sort mail efficiently. 

As mail volumes, along with the overall economy, began to boom after the war, the USPS turned careful attention to the logistics and logic of mail movement and separation. Zip codes were a natural development from the war-time experiment. The object was to make the USPS more efficient, and faster, in collecting, sorting, and delivering mail. 

Looking back from this vantage point in 2014 when a letter posted in Augusta, Maine in the afternoon on Tuesday will be delivered to an address in San Francisco on Wednesday, Thursday latest, the results were nothing short of astonishing. And all for 49 cents. And although it’s not 46 cents, it’s still quite a bargain.  Nowhere else in the world would this happen, for that price.

            The Zip code system plays a very key role in that progress. Mail can be directed optimally from any location to any location. It’s not for nothing that the USPS have one of the most robust and largest data processing capabilities of all government agencies.

            However, what sometimes gets lost in admiration for the Zip code system is that it was not intended to have the zip coded areas conform to anything other than the structure of the USPS delivery system, and that doesn’t always match city-town-county or even State lines. In short, the ZIP code system is based on mail volume, postal area size, geographic location and topography, but is not necessarily bound by municipal or community boundaries.

            And here is where Level Plains enters the picture, joining a fair number of other jurisdictions, by the way. Level Plains’ addresses, which the town uses to assess taxes and the like, are assigned through the ZIP system to the neighboring towns of Daleville and Enterprise.  Poor Level Plains doesn’t exist as a “postal data point”. 

            As a result, there is a great deal of “leakage” of tax revenue from Level Plains. Some of this is caused by modern accounting software, which applies tax and fee calculations according to Zip code, not according to street address, and that software is applying Daleville and Enterprise rates to Level Plains addresses. Some well-intentioned Level Plains tax payers are inadvertently applying the tax calculations of neighboring towns through incorrectly organized accounting software. And these rates are apparently lower than those of Level Plains. The solution of course is to give Level Plains its own Zip code and the USPS has a process to examine doing this.

            Why should this matter to marketers? Because your own customer and prospect data may inadvertently locate addresses in the correct Zip code but in an incorrect demographic or geographic location you may not want to address. If your client is a fast-food chain, for example, mailing solely by Zip codes may cause you to waste a lot of money. 

 Segmentation by postal walk would go a long way to solve that problem, of course.  With modern household data, an overlay and subsequent analysis of household characteristics on top of a Zip code map might show some surprises worth investigating. 

  In any event, to perform effective business intelligence yielding data analysis, and ultimately to get your mail delivered to its intended target, you’ll need properly standardized, accurate address data, with correct Zip codes, ideally down to Zip +4.  Look to Data Services to provide those domestic and international data quality and enhancement services critical to your success!