The Power of Accuracy

Finding Population Estimates You Can Count On

Executive Summary

In the science of marketplace analysis, wrong answers can upset even the best-laid business plans. Ever since companies began researching U.S. trade areas decades ago using geodemographic data and technologies, they have needed one variable above all to ensure high-quality market research - accurate population counts. But, for years, obtaining this dependable data has not always been easy - until STI: PopStats*™ launched in October 2001. This demographic product was created in direct response to researchers' urgent requests for population estimates that would give them greater confidence in their market research results and help them make more profitable location-oriented business decisions. This product has literally changed the way companies build trade area models and analyze demographic data. This white paper provides an overview of this innovative data product and compares it to other population-estimating methodologies used today.

Introduction

Population estimates. They are an essential component of trade area research, yet obtaining reliable population estimates has been an elusive goal for many companies. Businesses start looking for new locations by first looking for people who live and work in that area. When companies use inaccurate population estimates, their research suffers. When companies use correct estimates to analyze markets, their efforts yield successful projects, greater confidence, and smart investments.

One of the key problems in obtaining accurate population estimates stems from misunderstandings about the various methodologies used to create the estimates and their many pros and cons. Once the fundamental differences are understood, however, it becomes obvious which data product will give you access to the most accurate population estimates possible.

Another excellent way to learn which population data will give you superior results is by hearing directly from other market researchers how they obtained and used trust-worthy population estimates. Through their real-world experiences, market researchers illustrate how their market research activities improved and how they overcame specific problems - once they gained access to accurate population estimates.

This white paper delivers insights on both fronts by explaining the most common population estimating methodologies used today and sharing the experiences of a few of the many STI: PopStats users in the location-focused retail, real estate, and telecommunications industries. This white paper will show how using the most accurate population estimates available today will add considerable benefits to any population-dependent companies' market research activities.

State of the Market-Research Industry

How many people live and work in a trade area? The answer is the Holy Grail of trade area analysis.

By 2000, the market research industry had reached a new level of sophistication and researchers enjoyed access to a wide range of geodemographic data and technologies. However, researchers still could not dependably answer the essential question: How many potential customers live and work in this market?

Market researchers knew that inaccurate population estimates significantly compromised their research efforts by creating imprecise views of trade areas, weakening their competitive advantages, and eroding confidence in their answers. As a result, many companies were experiencing significant problems including (but are not limited to) the following:

  • Entering a new market prematurely, only to find that the projected population growth didn't materialize.
  • Failing to identify fast-growing areas and giving competitors the first-mover advantage.
  • Missing pockets of new high-growth neighborhoods inside trade areas.
  • Remodeling existing stores that didn't pay off due to slowing market.

Researchers also knew that when they conducted trade area analysis using accurate population estimates, they gained much more meaningful, confidence-boosting, and profit-generating research results. The advantages included:

  • More precise trade area insight.
  • Better business decisions.
  • Faster response to competitive advantages.
  • Greater confidence from the decision makers.

Over the years, companies have dealt with the problem of obtaining accurate population estimates in many ways. For example, they have compensated for the shortcomings by spending considerable time and effort to research markets through site visits. But, today, relying too heavily on this boots-on-the-ground approach or on time-consuming complex data investigations is often too costly and slows the research process down considerably.

Furthermore, today's tougher economic conditions do not allow room for the cost of research mistakes and expensive on-site visits. Yesterday's more cavalier attitude, which left a wider margin for error, is history. Now, companies cannot afford to invest in a traditional six-out-of-ten project success rate. Today, all projects must be profitable.

Understandably, by 2000 researchers were desperately searching the demographic industry for more reliable population estimates.

Three Methodologies - A Cautionary Tale

All population estimates are not created equal.

Until 2001, there were basically two methodologies used by demographic data providers to create population estimates: the "spread" method and the household direct marketing mailing and telemarketing list method. In 2001, Synergos Technologies introduce a new population-estimating product using an innovative and proprietary new methodology that brought zip+four postal delivery codes into the calculation. This section explains the differences in these three methodologies, and illustrates how easy it is to obtain population estimates that are either too low, too high - or just right.

SPREAD METHODOLOGY. This is the most commonly used approach to creating population estimates in U.S. markets. This methodology spreads, or allocates, population changes according to the historic rates of growth or decline across the U.S., states, counties, tracts, and block groups. But it has one serious flaw: It is based on historic growth rates established at the time of the 2000 U.S. Census. This growth ratio may or may not accurately represent actual population growth and declines rates in today's trade areas.

Simply stated, the spread methodology includes this basic process:

Estimates the U.S. population changes from one year to the next based on birth, death, and immigration rates. It then assigns percentages of the overall population changes to all states based on a combination of their population growth rates in 2000 and current-year information from the IRS, which helps to calculate residential migration changes from state to state.

  1. Takes each state's estimated population growth rate and spreads it across all of the state's counties based on their growth rates in 2000.
  2. Takes each county's estimated population growth rate and spreads it across all of the county's tracts based on their growth rates in 2000.
  3. Takes each tract's estimated population growth rate and spreads it across all of the tract's block groups based on their growth rates in 2000.

Due to four levels of calculation from the original population changes and a lack of IRS data past the county level, it's easy to see that there is a high risk of error in calculating population estimates using the spread methodology.

HOUSEHOLD DIRECT MAILING AND TELEMARKETING LIST METHODOLOGY. This methodology basically counts populations based on how many people are on certain direct mail mailing or telemarketing lists. There are two leading flaws with this process: (1) the consumer lists are typical one to two years out-of-date, and (2) during this time gap new people could have moved into a residence and been added to the lists, while former residents may not yet be removed from the lists. As a result, two or more residents with the same addresses could be on mailing lists. The problem with doubling (or even tripling) household population estimates is a serious issue when using these lists to calculate population estimates.

There are also two additional issues regarding the use of telemarketing lists in the calculation of population estimates. First, there is a long and growing list of U.S. residents who have put their telephone numbers on the national and state-based Do Not Call lists. Secondly, recent estimates show that a growing percent of the national population no longer owns landline telephones. In fact, by 2008, 40 percent of the population only owned cell phones. There are currently greater restrictions on the amount of telemarketing that can take place via cell phone. As a result, traditional telemarketing lists have lost a great deal of their potency in the past several years and have become an even weaker data resource for calculating population estimates.

STI: POPSTATS METHODOLOGY. STI: PopStats' methodology is completely different from the two traditional methodologies. For one thing, its primary source data is from the most current U.S. Postal Service data at the zip-plus-four level, which counts how many people are actually receiving mail at each zip-plus-four zip code across the country. Then it takes these numbers straight up to the block group level. This "bottom-up" approach allows population estimates to also be calculated straight up to any geographic region, including block point, tract, county, and state.

Each trade area's zip-plus-four levels are as small as specific groups of houses - typically four to 12 - or even just to a single building. As a result, accessing population data from this source allows researchers to literally see structures coming online as they are finished being built and occupied.

To further ensure accuracy, Synergos uses a series of checks-and-balances to validate he results, including consulting with multiple state and federal agencies whose data is independently gathered and calculated. Because this method works from the bottom-up and the controls come from entirely different and mutually exclusive sources, Synergos is able to provide the most accurate and unbiased estimates possible with STI: PopStats data.

There are two substantial benefits with the STI: PopStats methodology over traditional methodologies:

  1. There is only one level of calculation between the source data and the population estimates versus up to four with the spread methodology.
  2. It's based on people actually receiving U.S. mail according to the U.S. Postmasters' mail delivery data across the country, not according to mailing list companies, which frequently double or even triple count residents at each household address.

Accuracy in Action - Quotes from STI: PopStats Users

Real-world results offer the most compelling insight into the impact of accurate population estimates. When STI: PopStats data users are asked about the product, they cite many advantages, but one above all - accuracy.

Here are a few of the many success stories shared by STI: PopStats users.

Kroger Enjoys Greater Research Confidence. "We have been using PopStats since it first became available for three primary reasons. First, I was impressed that I could get all of my demographic data from one source. Secondly, PopStats was the only product that was updated quarterly. All of its data is the most current, including population counts, ethnicities, incomes, seasonality, and more. Third, I have great confidence in the source of the data, since it is based on residential postal delivery. Plus, the product's quarterly updates allow us to conduct new kinds of research, such as tracking changes in seasonal markets."
— Dale Caldwell, A.V.P., Kroger Corp. Development Research Department, Portland, OR

Weingarten Expands Development into New Markets. "To effectively expand into new high-growth markets, we needed regularly updated population estimates that accurately reflected current population changes. We originally accessed PopStats data to supplement our existing data from another provider. However, we became so confident with PopStats we decided to use its entire data set exclusively. We renewed our contract for semi-annual updates to save money. But it took only one quarter for us to realize our mistake - especially in today's volatile housing market. We immediately changed our contract back to quarterly PopStats updates. Now we are expanding new housing development with much greater confidence."
— Kyle Kretsinger, Director of Research, Weingarten Realty, Houston, TX

Family Dollar Gains Accurate Market Area Insight. "We have used PopStats since 2004 and have enjoyed greater confidence in our population counts, especially in market areas undergoing rapid change. Not only does PopStats give us accurate population numbers at the neighborhood level, but also it delivers insight on other important consumer demographic changes, including income levels, occupations, and ethnicity. With PopStats accurate quarterly data, we are able to build more powerful predictive models, assess information in the field, and make better real estate decisions - all with greater confidence than ever before."
— Brian Strickland, Director of Strategy and Business Development, Family Dollar, Charlotte, NC

Conclusion

Business decisions are fraught with risk. Even a slight error can wreck havoc on multiple fronts. Decisions related to business locations are particularly sensitive to errors. For example, even a one-block mistake in placing a new location can impact success or failure and profit or loss. In 2000, the market research industry had reached a new level of sophistication, but what it needed most was population estimates that companies could count on to provide the highest level of accuracy regarding who lived and worked in each market. This is exactly what they found in 2001 with the launch of a revolutionary new population-estimating product — STI: PopStats.

STI: Events

ICSC 2017 Research Connections
Oct. 28-30, 2017
Toronto, Canada

 


 

Mark your calendars for the coming:
STI: PopStats Research Conference
Mar 21-23, 2018
Austin, TX

 


 

ICSC RECon 2018
May 20-23, 2018
Las Vegas, NV