This year's User Conference was specifically created to help STI: PopStats users gain greater insight into our unique quarterly population-estimating tool and share other leading-edge market research tools, theories, and insights. This year our two-and-a-half day event was packed from sun up to sun down with presentations from Robert Welch, PopStats users, and industry experts. As a follow up to the event, we've prepared this User Conference Report with synopses of all the sessions. This is both a review of the highlights for those who attended and a guide to the ground covered at the event for those who could not be there with us.

As always this year's User Conference was an opportunity for everyone to share their thoughts, opinions, and experiences about market research with their peers. In keeping with our information exchange philosophy, we are also including a few of the many comments shared by the User Conference's attendees about each of the sessions.

TABLE OF CONTENTS

Wed. March 5, 2008
SESSION 1 - What's New and Major Initiatives Taking Place, Robert Welch
SESSION 2 - Marketing to the Wealthy Hispanics, Manuel Delgado
SESSION 3 - The Top 50 Growth Markets in the U.S., Robert Welch
SESSION 4 - Determining the Health of a Market, Danielle Jordan and Tom Kessler
SESSION 5 - Looking for Dollars, Tim Allen
SESSION 6 - Simulation Modeling, Robert Welch

Thur. March 6, 2008
SESSION 1 - How Understanding the Landscape Leads to Understanding Customers, Cindy Reed
SESSION 2 - Relating Lifestyles to Panel Data, the Saga Continues, John LeTourneux and Dale Caldwell
SESSION 3 - Retail GIS: Developing Smarter Systems, Mark Oster
SESSION 4 - Thinking Outside the Polygon: New Advancements in GIS/Display Technology, Craig Johnson
SESSION 5 - STI PopStats Methodology, Robert Welch

Fri. March 7, 2008
SESSION 1 - Defining a Trade Area, Robert Welch
SESSION 2 - Creating a Compelling Map, Robert Welch

Wed. March 5, 2008

SESSION 1 - What's New and Major Initiatives Taking Place
Presenter - Robert Welch, President, STI, Austin, Texas

What's New with PopStats? Robert reported that PopStats now contains 997 points of data, which is significantly more than when the product launched in October 2001 with 21 points of data. Since its inception on the market research scene as the first quarterly population estimate, the product has never missed a quarterly release. Its April 2008 release will be number 27. New features of this release will include:

• Seasonal data. This includes seasonal populations staying in a trade area for more than six weeks, but less than six months. Robert noted that you need to look at four quarters to see the normal seasonal fluctuations of any market.
• Transient data. The new aspect of this data is the histories of people staying in hotels, campgrounds, and RV sites for less than six weeks (the basic transient data was added in 2006).
• Unemployment data. This new information is based on data from the Bureau of Labor Statistics, and was added based on frequent requests for "econometric" data.

New data fields for upcoming releases:

• Subprime data. This will provide insight into today's expanding subprime credit sector and become part of PopStats's body of econometrics data. This will be available in the July or October 2008 release.
• Credit data. This will provide insight into the credit sector and become part of PopStats's body of econometrics data. The projected launch date is Winter 2008/2009. May be launched as a separate product.

New products being launched:

• STI: BlockPoint. This new add-on product provides population estimates at the block level for a finer-grained level of trade area research. Robert noted that while there are eight million block groups in the U.S., there are 28 million residential zip+four block points. The new BlockPoint product will be particularly vital as the country's large rural block groups become more populated. Typically populations only grow in a portion of these large rural trade areas. Now PopStats users can pinpoint the growth areas within large block groups. This product will launch in April 2008.
• STI: Zip Code. This new add-on product provides the same population estimate available at the block group level now at the Zip Code level, or more or more specifically at the ZCTA level.

Major STI Initiatives Underway:

• STI: ISIS. The Interactive Statistical Information System (ISIS) is a Map Info-based executive tool. Robert created it for internal use at STI. Now it will be available for PopStats users who want to be able to easily build customized reports. Technical support for ISIS is available through ROIC Analytics as a fee-based service. This tool is available now, upon request from any existing PopStats Licensee.
• STI: Data Depot. This new initiative will give PopStats Users access to a wealth of geodemographic data developed by Synergos Technologies, all for free, including Tiger files, USGS, satellite imagery links, and much more. The depot will be on-line within the next couple of months.

Attendee Comments on this Session Include:

• Good opening to the conference.
• This is always interesting and informative.
• Tons of useful info.
• BlockPoint data is very fascinating.
• This is a great intro - really liked this section.

SESSION 2 - Marketing to the Wealthy Hispanic
Presenter - Manuel Delgado, CEO, Agua Marketing, Houston, Texas

Manuel gave a deeply informative and highly entertaining look at the biases that often get in the way of properly marketing to the U.S.'s expanding Hispanic population - which currently stands at 44 million (more than Canada's entire population). This sector represents an $800 billion market. In fact, one half of the population owns their home. He said that one of the biggest problems with marketing to Hispanics is that "companies don't know what they don't know."

After revealing the common misconceptions, Manuel presented five powerful insights on the cultural realities that marketers must consider in order to effectively target this still growing market sector:

1. Who is the typical Hispanic? Realize that the common perception of Hispanics is only the average American's perspective. In other words, Ricky Ricardo, Ricky Martin, and Ricardo Montalban are not considered typical Hispanics by Hispanics.
2. Understand the culture. There are very specific keys to Hispanic culture, including the fact that family is critical, they came to the U.S. from particular circumstances, they hold specific religious traditions as sacred, and their language is important and should not be misinterpreted.
3. Be surgical in marketing. Marketers should understand where exactly Hispanics live and that among Hispanics there are cultural differences based on where they came from.
4. Understand the Hispanic decision-making process. Hispanics are likely to become loyal to brands they know either through association to what was available in their home country or by U.S. advertising.
5. Hispanics living in the U.S. are also Americans. Hispanics want to assimilate and become known as Hispanic Americans: Marketers should recognize this fact.

Attendee Comments on this Session Include:

• I enjoyed this presentation and you should look towards including more "lifestyle" sessions and profile ethnic trends of this type.
• Excellent! Great mix of comedy and fact. Very informative; best presentation I've seen in a long time.
• This was very insightful. I found myself agreeing with many of the misconceptions I had about the Hispanic consumer. This should be included in future conferences.
• Great talk, interesting topic, not relevant to my work. However, keep it coming because it's nice to see/hear something a little different.
• Excellent content and a very charismatic presenter.

SESSION 3 - The Top 50 Growth Markets in the U.S.
Presenter - Robert Welch, STI

"What a difference a year makes - especially this past year," stated Robert during his presentation of the current high-growth markets in the U.S. He pointed out that, despite the economy, markets are still growing; because there are still ~300,000 new residents per quarter based simply on birth versus death rates, which are "not impacted by the economy." However, migration has stalled with people "staying put" for now.

He pointed out that California, Texas, and Florida are still growth leaders. There is stable growth in the mid-Atlantic states. Also the over-50 population will begin to define the marketplace, gradually edging out the 30-year-old age group.

Robert presented a map and gave a short discussion on the high-growth markets - which actually only includes 49 markets this year (versus 54 in 2007).

Las Vegas was usurped from its long-held #1 position by New Orleans, and dropped to number 17. Robert said that from his observations the strongest growth in New Orleans is from middle America, "because they can afford to return." He also pointed out that the Carolinas took three of the 10 top spots on the list. PopStats's 10-year growth projections show Las Vegas retaking the number one position and Austin, PopStats's hometown, moving up to second-place.

Attendee Comments on this Session Included:

• This was a very interesting topic. I think that this gives a real perspective on growth in the U.S.
• Knowing how you did this helped to understand your results.
• Always a fun topic and interesting results.

SESSION 4 - Determining the Health of a Market
Presenter - Danielle Jordan, Managing Partner, ROIC Analytics, Boise, Idaho

One result of today's tighter economy is that businesses require greater "proof" that a new market is a viable new location. Danielle addressed critical questions that businesses must ask themselves including: How can market researchers bring this proof to the decision process? How can companies continue to grow in declining markets? What are the demographics that will make a difference?

Danielle pointed out a few particularly important market research questions to answer during challenging times including:

• How strong is the competition?
• Is there sufficient demand in the area for a new retail concept?
• What is the future expansion potential?

Danielle reviewed a case study of one of their clients where she proved to them using PopStats's data suite that relocating four stores in Boise would both slow the erosion of the company's profits and increase its market share. By comparing business scenarios using side-by-side data she was able to demonstrate that by doing nothing the retailer would only realize an 11.6% market share. But by relocating four sites, the company would boost market share to 21%. "This was impossible for the company to ignore," she says. "Due to this one decision, the retailer enjoyed a 50 percent increase in sales last year."

Attendee Comments on this Session Include:

• Extremely relevant for us right now.
• Great overview of market planning for beginner/intermediate/experienced individuals.
• Love seeing sample client projects. Would like to see more like this but with a bit more cutting-edge methodology.

SESSION 5 - Looking for Dollars
Presenter - Tim Allen, title, Building Committee, Inc., city, state

Tim presented a case study on the site analysis required for credit union customers, which he said, "provides a unique set of challenges." He showed how he built a matrix to identify the credit union's ideal customers and their locations in its trade area. Some of the findings from the research that impacted site selection included:

• Average user's preferred location used to be 50-50 between work and home proximity. Now it's 80 percent related to proximity to work sites.
• Despite increased consumer Internet use today, people still prefer to bank in person. As a result, there's little slow down in banks and credit unions opening new branch offices.

Attendee Comments on this Session Include:

• I like seeing samples of client engagements.

SESSION 6 - Simulation Modeling
Presenter - Robert Welch, STI

Attendees enjoyed a bit of humor as Robert worked to cut a particularly hard piece of Styrofoam in two. But the result was a greater understanding of the point he was making about slicing and dicing demographic data using simulation modeling. This session was created in response to feedback from the 2007 PopStats User Conference: Attendees' requested more insight on theoretical concepts in modeling populations.

Robert said the biggest value from simulation modeling is that it gives you the opportunity to make forecasts with uncertain data, such as market size, selling prices, market share, and more. It works best with ranges versus discreet numbers. Although it sounds difficult, Robert said it's "surprisingly easy" when you apply a Monte Carlo simulation model using uniform, normal, empirical, and triangular probability distributions. "You can make statistical inferences when you have a range," he added. The basic goal of the simulation model is to analyze the customers, not the stores, because "customers pick stores, the stores don't pick the customer."

Simply stated, simulation modeling includes these basic components:

• Develop a deterministic model.
• Identify uncertain variables.
• Quantify uncertainty with distribution.
• Analyze problems using a Monte Carlo simulation.
• Evaluate resulting distribution of outcomes (the result of having ranges of values gives you a distribution).

Attendee Comments on this Session Include:

• As usual, it was great. I love the theory. Do one each day!
• Need to continue - it stretches us - that is good! Please continue.
• I think theoretical discussions of this nature are very beneficial as a "counter-balance" with the real-life discussions of other presentations earlier.
• Awesome! Loved it. Keep the "theoretical" stuff coming.
• This presentation helped me understand the model I use in my current job. Very insightful and should be touched on every year.
• Love to hear the theory-I probably would not attempt but it sure makes me look smarter!

Thur. March 6, 2008

SESSION 1 - How Understanding the Landscape Leads to Understanding Customers
Presenter - Cindy Reed, title, Mapping Analytics, Rochester, New York

Cindy presented a discussion on maximizing market research with STI: Landscape. "Essentially the best research is the study of differences. The goal is to find out what is different between our customers and all possible customers. We are looking for clusters where there are differences, then we are measuring the differences between consumer groups."

From her experience, Cindy said that many researchers tend to think that segmentation is just demographics. But it is "also attitudinal. The more information measured by your segmentation system, the better it will discriminate between consumer groups. That's why Landscape's attitudinal measurements are so valuable. They add an extra layer of detail for comparison."

Also it's critical that market researchers "identify exactly who their customers are versus who they want them to be. First find your base and build your business around your largest cluster of customers. Then you can target customers you would like to attract with a whole new marketing program."

Attendee Comments on this Session Include:

• Very useful, great teaching.
• We would like to hear more from her.
• Great subject and presenter.
• Very enlightening.
• Cindy is a great presenter, thorough and informative.

SESSION 2 - Relating Lifestyles to Panel Data, the Saga Continues
Presenter - John LeTourneux, Senior Analyst, The Kroger Company, Portland, Oregon

This session was part three in an ongoing discussion from Kroger started at the first annual PopStats User Conference. This year, John described how the largest U.S. grocery chain (by units) is attempting to tackle the challenge of relating Landscape lifestyle data to panel survey data. The company's goal is to gain customer insight to make smarter decisions on redesigning existing stores. With X stores in X markets it takes sophisticated research to segment its customers in each trade area.

"Can we become more intelligent by combining value card customer data with our Landscape data?" That's the question that is driving Kroger's current research - and its constantly evolving mission to conduct more result-oriented market research. Eight-five percent of Kroger's customers use value cards, which allows the company to track what people purchase, how much they spend, and more.

The process starts when the GIS department is given a group of stores with problems. To find solutions the researchers start by dividing them into two categories: Those with image problems and those with product selection problems. Kroger has found that there are 412 products that are the key indicators for its ideal customers, which are determined nationally versus regionally. They then compare the national norm to what each store actually sells. "We are looking at the movement of product based on the averages," noted John. "We need to know the bad and good to make the best decisions."

Kroger is hoping that combining Landscape data with actual sales data will help the grocer make more profitable decisions on store changes. For the results of Krogers' current innovative demographic project, be sure to attend the 2009 PopStats User Conference.

Attendee Comments on this Session Include:

• John did a super job in keeping his presentation simple and easy to understand the concept. The presentation wasn't confused with too much detail.
• Great application of the data.
• Good real-life example as to what a major retailer is going with its tools.
• Great to see another application of Landscape.
• I enjoy the way the guys from Kroger explain their long-term projects and appreciate their willingness to share.

SESSION 3 - Retail GIS: Developing Smarter Systems

Presenter - Mark Oster, title, Oster Research, city, st

In Mark's talk, he explored the question: Can we use GIS as an "expert" system without hitting the typical "human research speed bump" along the way? He said that today's typical GIS process requires experts who are the knowledge holders of all of the retail information. They study all of the data and provide their opinions based on the numbers. The problems with this traditional approach includes:

• GIS plays a sub-optimal role in the process and cannot produce "actionable answers."
• Analysts do too much heavy lifting to get to the answers and can miss the big picture.
• Market studies become too long, time-consuming, and ineffective.

An alternative is to bring "formal retail knowledge" into the GIS process using hypothesis testing that incorporates a lot of data. The GIS analysts only need to insert their informal retail knowledge, which lessens the time and work required to complete trade area studies. They are effectively freed from "recalling anecdotes and pouring over minutia." They can quickly and easily produce reports that are concise and easier to read, and which express critical information in "development terms," not GIS terms.

Mark gave an overview of his consulting company's four-step GIS process used to study the retail opportunities in a shopping center. He gave specific examples for a retail developer who used the "market potential" reports to attract retailers. The steps include:

1. Demographic indexing. Critical measures, such as population, growth, and incomes, are indexed to existing stores and compared with the universe of relevant locations.
2. Development opportunity. This determines what the development opportunity is, who would be suitable tenants, and if the demand out weights the supply, if the supply out weights the demand, or if they are equal.
3. Tenant suitability. This step identifies which retailers are the most suitable tenants for a location based on geodemographics.
4. Supply-demand analogs. The final step makes site opportunity comparisons among the analogous retail options.

Attendee Comments on this Session Include:

• Really enjoyed this type of presentation.
• I found Mark's presentation to be very fascinating and useful. I liked seeing how Mark integrated Google tools with other data to improve upon his presentation. Adds a "wow" factor to presentations to potential customers.
• Appreciate the cutting-edge subject.

SESSION 4 - Thinking Outside the Polygon: New Advancements in GIS/Display Technology
Presenter - Craig Johnson, Managing Partner, Geographic Enterprises, Boulder, Colorado

Craig explained and demonstrated a new location-modeling technology called DERM, the Digital Earth Reference Model. It is an innovative way to extract geographic information data from every location on the earth using a polygon shape. He explained that the military is using the technology because it delivers a wide range of advantages that are equally valuable to the retail sector including:

• Overcoming inconsistencies in mapping data.
• Handling massive amounts of data.
• Rapidly seeking insights and formulating decisions based on the data.

The technology's mission is to "connect vast amounts of disparate data, one of the last great needs of modern data research," noted Craig. The purpose of the technology is to "make data easy to interact with and to see visually. Our tools allow researchers to connect to a wide variety of databases, from workstations to corporate, so they can rapidly deal with millions of records."

Attendee Comments on this Session Include:

• Great info to stay on the cutting edge.
• I found Craig's presentation to be very interesting and useful in that it intrigued me as to the alternative theories and technologies.
• I like a presentation of future or "leading edge" technologies and applications.
• Wow! This could be the future.
• Very interesting topic. I would like to see this topic expanded on in future conferences.
• Hexagonal data could be a huge development in the world of GIS/demos; well presented.

SESSION 5 - STI PopStats Methodology
Presenter - Robert Welch, STI

One of the most popular sessions at all three PopStats User Conferences is Robert's description of PopStats's methodology. If you missed the presentation, be sure to attend the event in 2009. Here is a brief description of the methodology:

These are the many models that comprise the STI: PopStats's model:

• 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.
• 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+4's 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 (if-then questions). In short, it is the final decision-maker of the estimate.

The STI: 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."

Attendee Comments on this Session Include:

• Need to know and find it very relevant.
• Even if you've seen this, it is good to see it again to refresh one's understanding of methodology. Quite often I get asked about "how" I got my answers. This session is where I get my answers to that question.
• Excellent presentation for this first-timer!
• Need to have notepaper for this section. Good stuff!
• Best session of the day.
• This helps customers have even greater confidence in the product.
• Core value to the conference.
• This is one of the reasons that I am here.
• Understanding the data helps you feel confident about using it.
• Very informative. Main reason for attending the conference.

Fri. March 7, 2008

SESSION 1 - Defining a Trade Area
Presenter - Robert Welch, STI

Robert reviewed some of his own methods for defining trade areas, which he termed "Robert's Bag of Tricks," including:

• Pie study. This is a variation on the ring study. While ring studies typically provide a macro view of a trade area, a pie study provides a smaller, and therefore more specific, area to research. While ring studies tend to average out the area, a pie study can target more narrowly. Robert noted that this is particularly valuable for large market areas.
• Map tiles. Unlike trade areas, which represent only a specific area, tiling can provide a much better representation of a particular market, including all of the many areas from which it could be pulling customers.
• Customer density. Robert said that one innovative way to do this is using hex grids, which "make great circles. People tend to think in circles, so hex grids are a logical shape that overcomes some of the issues presented by circles." (He added that Craig Johnson's company may be on the cutting-edge of this development.)
• Barrier composition. This study is related to PopStats's new BlockPoint product. It involves creating trade areas bounded by geographic features, such as highways and waterways. This provides a naturally occurring trade area, since people are less likely to cross geographic boundaries to shop.

Attendee Comments on this Session Include:

• Out of the box thinking.
• Very informative; interesting.
• Terrific.
• The additional programs/hints very helpful. Thanks.
• It is very interesting to see Robert take this on. He has some great methods as compared to some more specific of what a user would present.
• I really liked the detail.
• These presentations were some of my favorites and have such practical, immediate application.
• This was very useful. Can't wait to try Map Bar.
• Cool segment.
• Great session.

SESSION 2 - Creating a Compelling Map
Presenter - Robert Welch, STI

"Mapping is just telling a story," said Robert at the start of his talk on creating compelling maps. "You are trying to get across specific points, so you need your maps to clearly communicate your point." He said that the four basic elements of maps are:

• Platform. The platform, either paper or electronic, significantly impacts the creation of maps and their elements.
• Color. Colors help to focus the eye and direct viewers to the issues you want to communicate.
• Texture. Texture not only brings life to a map, but also introduces subtleties they may be important to a business decision. For example a mountain range or gorge.
• Content. Of course content is the reason for being. Robert introduced a number of techniques for driving different points home.

Attendee Comments on this Session Include:

• This was very helpful. For those of us who have used GIS maps for a long time, it was interesting to see truly innovative new display techniques.
• My favorite topic. Very well done.
• I think our dept. makes great maps already, but we gained some insight and tools to significantly enhance them. Thanks.
• This portion showed how limited our current map software is.
• It's true that people do not care about the map, which is wrong. I really learned a few tricks to come up with a good map.
• It was a nice review on some cartographic concepts and I love the tools Robert creates.