New-unit Site Selection and the Use of Quantitative Modeling Around Guest Demographics and Psychographics
By Todd Walter President & Chief Executive Officer, WTS International | March 04, 2012
"Where to open next?" - that is the $64,000 question every retailer asks itself as it pursues growth. Whether you're selling products or services, the first three rules of real estate remain: location, location, location. But how to identify and prioritize those locations to maximize your probability of success, particularly in a difficult economic environment, is the question that matters most.
This is the question we at Red Door Spas faced as we looked to expand our geographic footprint across North America. In late 2006, our senior management team identified 21 major metropolitan markets, or "gateway cities," that we felt we should be in. At the time, Red Door Spas had locations in just six of the 21. But even with the markets identified, we then faced the challenge of prioritizing those markets and finding specific sites within them.
In addition, as a private-equity-backed company, we wanted to be able to articulate our expansion strategy in a way that not only made sense, but that could be easily understood by our owners and other financial constituents. For help, we turned to Site Analytics Co., a real-estate services firm that has been helping retailers and developers for the past 15 years.
The first step in our process was to develop a thorough understanding of our existing spas' performance and to identify critical success factors that contribute to their relative financial performance. This analysis involved both site-specific considerations (e.g., local demographics and psychographics, host location attributes, co-tenants, and competition) as well as Red Door-specific considerations (e.g., physical space attributes, staff turnover, and manager tenure).
Through regression analysis, over 1,000 variables within these and other categories were tested for their predictive value in determining a spa's actual, historical revenue. A revenue forecasting model was then built using those variables showing the greatest predictive influence. For the statistics junkies out there, our model's coefficient of determination, or R2, is .78 (an R2 value of 1.0 represents perfect correlation).
With our predictive model in hand, we then mapped out all 21 North American metropolitan markets and forecast by one-square-mile areas what we should expect a Red Door Spa to generate in revenues if situated in each area. In total, we have identified approximately 125 target areas that could contain a Red Door Spa forecasted to perform at levels similar to our most successful Red Door Spas. Next, we prioritized gateway cities based on their respective total market potential (i.e., how many successful Red Door Spas the model tells us the market can sustain) and further, we prioritized areas within those cities based on the predictive model (i.e., pursuing first those areas that are expected to yield the highest revenue per spa).