2006 PopStats Research Conference Summary

STI: PopStats' 2006 User Conference a Success!

The first annual STI: PopStats Research Conference and Forum was held on March 23 and 24, 2006, in Austin, Texas. Our Users Conference was specifically created to help STI: PopStats users gain greater insight into this powerful, one-of-a-kind quarterly population estimating tool. And we're happy to report that we delivered on this mission. At the event, the attendees shared two full days of insightful presentations given by industry leaders, peer-to-peer networking, and product development brainstorming.

On the first morning of the conference, Robert Welch welcomed the attendees and said STI: PopStats' mission is to continue to deliver the highest level of demographic value to today's retail market researchers. STI strives to fulfill this mission by staying on the leading edge of market research. And the best way to stay cutting-edge is to stay connected to the product's users.

For those who were unable to attend, we've created this overview of the insightful information delivered by the event's speakers.

Table of Contents

1. Demographics: Where it's Been. Where it's Going.
2. STI: PopStats' Methodology Overview
3. Risk, Retail, and Katrina - Modeling After a Disaster
4. Wall Street Meets STI: PopStats
5. Data in Motion with TerraSeer
6. Walgreen's Targets Retail Locations Using GIS
7. Kroger's Trade Area by Design

1. Demographics: Where it's Been. Where it's Going.

PRESENTER: Larry Maves, Recently Retired MapInfo® Executive and 30-Year Demographic Industry Veteran

Larry Maves could be called the "father of demographics" for his long and influential career in the industry. STI was honored to have him come to the first Users Conference to share his singular view of the industry's history and future. Larry led off his discussion with this pronouncement: "STI: PopStats is a forward-thinking product."

Larry acknowledged that the primary reason that market researchers attended this event was to learn how to get better answers from their data and accelerate their answers about questions such as: Where are our customers?, Where are high-potential new markets?, and What are viable new retail concepts? In the pursuit of the answers, time is of the essence. As such, every researcher needs to create fast GIS systems loaded with relevant data.

The data that researchers use to facilitate their queries is critical. Plus, understanding data's capabilities and shortcomings is fundamental to successful research. Larry recommended that every retailer build data warehouses with a wide range of high-value data from four areas:

1. Externally acquired data (static) - including demographics, employment and business counts, consumer expenditures, panel survey information, traffic volumes, and competition data
2. Client transaction data (dynamic) - including sales, merchandising, POS transactions, marketing and advertising, and warranties
3. Client documentation data (static) - including leases, title policies, contractor and vendor data, audio-video information, CAD drawings, and franchise agreements
4. Client customer data (dynamic) - including location, householder transactions, high frequency and high volume data, lifestyle profiles, also-bought product data, and media response transactions

Larry cited STI: PopStats as the best tool available for delivering demographic data that is highly accurate, timely, and useable. "It's light years beyond what we used to gather in the field," he said. Because of STI: PopStats' unique quarterly updates, Larry pointed out that completely new variations in retail performance can be assessed - in other words, assessments that were never before available to retailers are now available.

Also important is the process that researchers use to extract relevant answers from the millions of data points available today. "We continually face the issue of filtering down to the data that's most important and getting as much insight as we can, because as we all know 90% of the research process is interpreting the data," Larry said. "Then we have to get it into the hands of the users quickly, so it has relevance. Six or seven months later is not an option."

The good news regarding accessing these data-rich data warehouses is that the tools to extract the most relevant data are maturing, he noted. These tools are allowing retails to access relevant data and extract it in ways that are meaningful to many areas of the business, including merchandising, human resources, finance, marketing, real estate, and operations.

Larry also spoke to the evolution of consumer lifestyle segmentation during his years in the industry. It's not just valuable to seek answers to the question "who." For the most profitable answers, researchers also need answers to what, where, when, why, how, and how much. He said that as segmentation systems have matured they have gained the ability to answers these critical questions. He pointed to STI: LandScape as a premium example of consumer segmentation. "It's the equivalent of hunting to understanding our customers' wants, desires, and changing needs with a rifle and scope versus a shotgun," he said.

Larry said that he's seen many retailers who will not invest in advanced data products. "Yes, good primary data is expensive. But it's not as expensive as being wrong when you're making strategic market decisions that can help expand your company's success."

So what's a cost-conscious retailer to do? Use the data with a sense of purpose, suggested Larry. Start building valuable databases, filter down to the salient points that allow decisions to be made rapidly, and get it into the hands of the people who can use it. "To spend your research dollars wisely, understand when to use the data and when not to use it. When it will bring value to the decisions and when it won't."

Understanding data's capabilities also allows researchers to expand its use into new areas. "Play with it and take it further than you are today. Never believe you've conquered the mountain. Keep looking for better answers and deeper insight. Ask yourself: how do I take what I've got and do it better? It's a simple concept, but very powerful when practiced."

2. STI: PopStats' Methodology Overview

PRESENTER: Robert Welch, President, Synergos Technologies

"How was STI: PopStats created?" This is the most frequent question Robert Welch is asked by market researchers. For the first time ever, he described the unique methodology Synergos Technologies used to create STI: PopStats' quarterly population estimates.

"Before we created PopStats we were creating software and data products. But invariably every market researcher who arrived at our doorstep had one similar request: `Where can I get accurate population counts?' We searched the demographic world high and low and found nothing to solve this fundamental problem. So we decided to create our own population estimating methodology. And instead of looking at the way other companies were creating population estimates, we decided to start from scratch by asking ourselves: How would we do it?"

Robert delivered a pragmatic discourse on PopStats' methodology beginning with an overview of the many models that comprise the PopStats model:

• The ZIP+4 Model. This model is based on over 28 million ZIP+4s, representing over 116 million households. This model is the primary determinate in understanding population growth and decline. Vital to the process is that Synergos Technologies maintains its own street files that feed into PopStats, because "we do not want spurious third-party data entering into the calculation."
• Postal Delivery Model. This model is based on postal delivery statistics provided by the postmasters in each market across the country. This model's primary purpose is to understand trends in existing populated areas.
• Spread Model. This model, which is based on macro-level postal counts, performs double-duty: it both calculates populations in rural areas where ZIP+4s are limited, and serves as a checks-and-balance for the previous two models.
• Census Model. This is the grand master of all the models, which STI also calls the "black box." It pulls together the other three models using an extreme set of heuristics (for example, if-then questions). In short, it is the final decision-maker in the estimate.

The PopStats methodology also includes automated processes for overcoming any and all anomalies present in the data, including ZIP+4 inaccuracies, data smoothing issues, conversions (lofts), and overrides. Robert said that the STI: PopStats estimates are calculated on six computers working together in a pseudo-parallel processing manner. "We have created a self-correcting artificial intelligence modeling system that learns from itself."

Should a client wish to challenge PopStats' estimates, Robert told the conference attendees that there are two levels of responsibility: Synergos Technologies' and the clients'.

1. Synergos Technologies' responsibilities include verifying the issue presented by the client, check the underlying data, checking soft sources, and reporting back to the client in a timely manner.
2. The client's responsibility is to gather as much detail as possible, provide Synergos Technologies with any third-party evidence that supports the client's position, and be patient during the evaluation process.

3. Risk, Retail, and Katrina - Modeling After a Disaster

PRESENTER: Robert Welch, President, Synergos Technologies

Hurricane Katrina was like no other disaster to hit the U.S. In this session, Robert Welch described how STI achieved the seemingly impossible task of determining a sound population estimate for the markets impacted by Katrina, which other providers said could not be done!

To create a model to calculation populations following the Katrina disaster, STI first:

• Separated fact from fiction
• Separated the temporary from the permanent
• Ignored the traditional data sources that proved unreliable
• Started thinking creatively to find new, undiscovered data sources

Among the new sources of data that STI identified were local newspapers, FEMA, USGS, Army COE, local school systems, and changes of address from the USPS. The final conclusion on who lost and who gained populations was that only three parishes and three counties lost significant populations, and approximately 500,000 people relocated to 500 U.S. counties.

Read all of the details about how STI created a post-Katrina population estimating model in the first issue of STI: PopViews Q1-2006. The only significant update to this detailed overview of modeling after a disaster is that STI took a second site visit to the New Orleans area in February 2006 to calibrate the updated population estimates in the latest release of STI: PopStats on April 1st, 2006.

4. Wall Street Meets STI: PopStats

PRESENTER: Abhay Padgaonkar, President, Innovative Solutions Consulting

Abhay Padgaonkar has known Robert Welch since they worked together in the 1980s. He recently conceived of a way to apply STI: PopStats to retail stores' market research activities and achieve more insightful and accurate results. He described this innovative process and its benefits to the User Conference attendees -- and asked if any company was interested in being a beta site for this breakthrough concept -- which he calls "Storefront." The following is an overview of how Storefront will work.

Today retailers measure their stores' performance using a range of metrics including:

• Relative to itself over time
• Relative to other comparable stores
• Relative to competition
• Relative to the external environment

But identifying which stores are the best and worst performers is not always obvious.

Storefront leverages quarterly population estimating demographics to give retailers a powerful new way to measure their stores. By adding this data to the equation, they can correlate store performance to population benchmarks to rank-order stores and answer these critical questions:

• How has the store or network performed relative to population changes?
• How effective are marketing and advertising strategies?
• What is competitions' impact?

Abhay explained how this information can be used to improve sales forecasting, new site selection, and staffing/operational efficiency -- allowing retailers to instantly see if store sales outperformed population changes and by how much.

To analyze store performance, Storefront will give retailers the following capabilities:

• Each store will have a data record with the internal store id number, its location, quarterly sales for the last eight quarters, and other attributes such as the size, format, tenure, and competition
• The ability to input and import store performance data from Excel/dBase into Storefront
• Plotting the information for a selected group of stores

To analyze customer and trade area information, Storefront will give retailers the following capabilities:

• Each customer will have a data record with the internal customer id number, location, gender, age, income, annual store spend, lifestyle segment, and household size
• The ability to input or import customer and trade area information from Excel/dBase customer information into the product
• Create trading areas using a variety of different methods, including fixed, radius, 80% of customers, drive time, and a huff model

To analyze customer attributes, Storefront will give retailers the following capabilities:

• Access the customer information within the trading area for a given store via a visual color-coded chart
• Each color could represent a different value of a selected attribute, such as age, income, or annual spending, for easy and fast analysis

To analyze population changes within trade areas, Storefront will give retailers the following capabilities:

• Once the trading area is nailed down, retailers would be able to see visually how the overall population has changed over time in the defined trading area
• Retailers could also see how the attribute mix has changed (income, age, gender, race, etc.), or how the population of specific neighborhood segments has evolved within a specific trading area

In the conclusion of this presentation, Abhay outlined the many powerful advantages available to retailers from this data analysis process including:

• Performance management - assess network/store sales, coverage, penetration, share, and profitability performance
• Marketing effectiveness - judge the effectiveness of target marketing and advertising strategies
• Competitive analysis - determine the impact of competition
• Business analysis - conduct sophisticated analysis (e.g., actual vs. potential using STI: Spending Patterns or population/customer attributes vs. specific merchandise using STI: LandScape)

The insight delivered through Storefront will allow retailers to achieve two strategic objectives:

1. Process improvements - identify opportunities for improving sales, coverage, penetration, market share, and profitability
2. Decision making - enable better sales forecasting, new site selection, and staffing/operational efficiencies

5. Data in Motion with TerraSeer

PRESENTER: Nicholas Jacquez, President, TerraSeer

As every GIS pro knows, compiling relevant geodemographic insight is one thing and sharing it with end users is another. Nicholas Jacquez, President of TerraSeer, demonstrated an innovative new way for market researchers to share market knowledge with decision makers -- using a time-series intelligence system to visually animate changes in geodemographic data. "Business analytics, though powerful, are meaningless unless decision makers can easily interact with the information they need," said Nick. He then described this unique data modeling product and its benefits to the Conference attendees.

"The convergence of spatial information and geographic technologies with Business Intelligence (BI) tools is an exciting trend, but two key aspects need further development if geodemographics is going to move into the mainstream," noted Nick.

1. Geographic technologies help the user visualize and model the marketplace, but events in the marketplace occur in geographic space and time. Therefore, to optimize results, market research inputs must examine "when" as well as "where."
2. As the pace of business quickens and analytic complexity grows, Retail Business Intelligence tools must be quickly understood, highly visual, and integrated into easy-to-use work-flows, so that line managers, not just analysts, can use the information to make accurate decisions, rapidly.

Especially for retail, BI has not provided business leaders with geographically enabled tools that easily support the retail management style. The dynamic, fast-paced environment of retail extends from the individual store to the market and regional level and all the way to the executive suites. Every day, retailers confront high-impact problems that require immediate resolution. The decisions must be based on the best information and analysis available, and formatted to support the intuitive decision style native to many retail business leaders. The key opportunity for Retail Business Intelligence is to provide a platform that allows managers to:

• Visualize their business networks in a dynamic and highly communicative interface, and
• Reveal key relationships in a way that fosters intuitive and fact-based decision making, without needing analysis expertise.

The TerraSeer Space-Time Intelligence System (STIS) is the first true space-time information system, Nick explained. The TerraSeer STIS is not a GIS, nor is it based on GIS technology. In the STIS, time is an integral part of the data and the software interface. As a result, all views of the data can be animated, from maps to tables, giving users a visual understanding of data changes through space and time, and an opportunity to interact with the data in real-time. Plus, STIS does more than just animate data; it also links selected data views together and helps make statistical inference regarding patterns in the data. The end result is a powerful new tool that supports dynamic business decision-making.

Nick summarized the features and benefits of the TerraSeer STIS for retailers:

• Like a GIS, the STIS enables mapping, querying, and graph making. Unlike a GIS, the STIS integrates time into the data structure, so maps, queries and graphs "move," letting the user visualize data changes through time.
• Allows on-the-fly animation of points and polygons -- so retailers can watch people, places, and situations change over time.
• Links windows to allow data exploration -- promotes rapid, intuitive insight as events unfold.
• Multiple data views (maps, graphs, and tables) are linked so that features highlighted in one view are simultaneously highlighted in all other views.
• Built-in spatial and spatio-temporal statistical tools let users see change over time and test its significance, without having to export for analysis to statistical software.
• Fully GIS compatible, STIS can be used with a GIS or on its own. Export formats are STIS project files and shape files.
• Flexible and customizable through API - it's easily integrated with existing workflow.
• Accepts industry standard data formats - it enhances current technologies, without disruption.
• Versatile platform technology - broad applicability, including retail, HL security, marketing, and business intelligence.

6. Walgreens Targets Retail Locations Using GIS

PRESENTERS: Dave Miller and Jillian Beydilli, Walgreens GIS Team Members

In the "retail war," Walgreens said that STI: PopStats is a key weapon in its arsenal for selecting the best retail locations across the country. Walgreens is among the most aggressively growing U.S. retailers today, employing continually evolving GIS processes to analyze new and existing markets. But behind the scenes, the drug store chain's market researchers have faced many challenges -- many of which they've conquered with their progressive application of trade area drawing and the application of STI: PopStats.

"Our aim has been to lower the risk of opening new stores and entering new markets," said Dave Miller. "Achieving this required getting more sophisticated with our market analysis and advanced tools that streamlined our ability to define and draw trade areas. And once we had the advanced tools, we realized we needed better data -- including accurate population counts. That's when we found STI: PopStats. It's one of the tools that has help generate information that we have confidence in -- without going into the markets on personal site visits and counting heads by hand."

STI: PopStats has been particularly critical in achieving the company's challenging growth objective: being the first one to the corner, said Jillian Beydilli. "We don't want to open new stores too soon or too late. GIS is the critical factor in making the right decision. With more advanced tools, we can achieve better answers." Walgreen's GIS team proactively passes its insight onto key departments, including sales, marketing, advertising, and category managers.

NOTE: Due to corporate policy, Walgreen's presentation cannot be reproduced in detail.

7. Trade Area by Design

PRESENTERS: Dale Caldwell, AVP, and Karie Hemming, Analyst, Corporate Development Research Department, Kroger Company

Imagine trying to define over 2,500 unique trade areas based on hundreds of thousands of customer spottings in just a few weeks. Kroger's GIS team managed this task by creating a unique modeling project and employing intelligent demographic data -- including STI: PopStats. Dale Caldwell and Karie Hemming described how they did it at the STI: PopStats User Conference.

Kroger's insightful session detailed how the grocery store chain uses STI: PopStats to make more informed trade area decisions regarding selecting which stores to remodel for the highest ROI. This is just one of Kroger's applications of STI: PopStats in the grocery store retailer's ongoing highly sophisticated geodemographic research.

While previously the corporation made store-remodeling decisions based on non-demographic reasons, such as store age and dollars earned, now the chain applies demographic research to make the decisions. "The answers are not always obvious, and at $3 million per renovation, it's an expensive mistake to make," stated Dale. The GIS team created its own unique program, which they call the Priority Index Model project (PIM), for increasing the effectiveness and efficiency of applying demographics to this problem. "PIM helps us keep our priorities in focus and ensures we achieve a greater bang for our buck," noted Dale.

The objectives of the project are to provide both the projected five-year population growth percent and median household income for 2,510 stores based on their individual trade areas. Dale and Karie explained how PIM helps them create more accurate trade area boundaries and reduce their research time per store from 15 minutes to two minutes. "Achieving these efficiencies made the project feasible," said Dale.

Simply stated, Kroger's PIM program uses STI: PopStats and the parameters of distance and density to create a spreadsheet that rank orders stores according to their growth rate and incomes. The higher the demographic growth rates and incomes, the more points a store receives. "The higher scoring stores float to the top, so we know which stores will deliver higher ROIs from remodeling investments," explained Dale.

Kroger's GIS team divided the project into three segments:

1. Creating a trade area around each store. They captured 70 to 80 percent of customer spottings to create boundaries for all stores by one of two ways: (a) for stores with loyalty card programs, they extracted customer addresses, then sorted, separated, and formatted the addresses using Access and Excel, and geo-coded the results; or (b) in stores without loyalty cards, they drew a three-mile radius.
2. Extracting demographics from STI: PopStats. This included creating a trade area file, inputting information into the file, and saving the file as a dbf.
3. Formatting information into an Excel spreadsheet. They formatted the demographic data from STI: PopStats into Excel worksheets using an Excel macro to calculated five-year population growth rates. Then they sent the worksheets to all 17 Kroger divisions.

Every six months Kroger updates their PIM analysis. "Thanks to STI: PopStats' updated population counts, we can access current insight on our stores' demographics and deliver up-to-date remodeling recommendations to Kroger's corporate offices."

What's Ahead for the STI: PopStats 2007 User Conference?

If you missed this year's STI: PopStats User Forum and Conference, make sure to mark your calendar for next year's event when we announce the dates later this year. We don't want you to miss your chance to gain unprecedented knowledge during two full days of insightful presentations given by cutting-edge industry leaders, networking with your industry peers, and share your thoughts on STI: PopStats future product development.

We're planning to expand our attendance capacity next year, so that even more of today's most progressive market researchers can gain the unique market research delivered in this one-of-a-kind industry event. Next year's conference will again be held in Austin -- so also plan to extend your stay for as long as possible so that you have time to research the many unique retail developments underway in the booming Silicon Hills.

Announcements