Retail site selection is not just a matter of available real estate. It’s an analytical challenge that requires sophisticated statistical modeling and a thorough understanding of the market potential of a location. Devon Wolfe, managing director of Americas Strategy & Analytics, Troy, N.Y.- based Pitney Bowes Business Insight, discusses the Top 10 most common mistakes that seem to repeatedly plague businesses and hinder their success:
1. Assuming you can build a model for all situations
Retailers often make the mistake of trying to develop a “one size fits all” mold for their new location, but the key to good store performance modeling is to know when a separate model is necessary. As an example, many chains have stores in a variety of shopping types: regional malls, strip centers, and street-front locations. Developing separate models for each of these location types allows the retailer to account for the specific locational factors that are present in each situation.
2. Failing to understand the limitations of a model
Retail models cannot directly replicate situations that aren’t present in a database of stores that already exist. For example, a retail concept with an entirely suburban real estate strategy will be hard pressed to accurately model the first entry into an inner-city market. It is unsafe to assume that a model will be able to adjust for atypical situations.
3. Failing to verify store location
While rapidly improving technology means tools like commercial geocoders become more accurate with every release, it is not uncommon to find that -- due to factors such as address errors -- a geocoder returns a location that is ¼ mile, ½ mile or even several miles from the true location. Yet, many companies continue to make multi-million dollar decisions based on these technologies without any further verification. While it may be time consuming, physically verifying your site’s location can make an enormous difference in the results you get from your predictions.
4. Using inadequate or undersized samples
Many retailers either have no data at all on their customers or they have partial data such as a limited-period customer survey, a database of in-store returns, comment cards, mailing lists, etc. Although some of these databases may be representative of the overall customer database, some may reflect a significant bias. Developing a customer database that is adequate in size and free from bias can prevent retailers from being mislead into making costly commercial property mistakes.
5. Using inappropriate variables
Retailers should remember that statistical techniques like correlation measure association, not causation. Also, variables that describe a very small portion of the population are much more likely to produce false correlations than variables that are well represented in the population. Use the common sense check -- if it doesn’t make sense that pet ownership and Rolls Royce ownership should be highly related, it’s probably not a good idea to use that relationship in a model.
6. Overfitting statistical models
While there are many complex factors that influence sales performance, many retailers may not recognize that overfitting a statistical model can do more harm than good. A model with a certain number of variables may do a good job of explaining the data from which it was drawn, but would likely do poorly in new situations.
7. Inadequately measuring competition
Many models in the location industry look only at the radius counts of competition without considering how store performance is affected by the positioning of competition in the catchment. Some retail concepts benefit from nearby competition; others don’t. For example, a fashion boutique would benefit from a location close to similar stores that will attract recreational shoppers. A model that hasn’t considered such issues has not adequately addressed the competition.
8. Using the wrong base for the customer profile
Segmentation systems can be powerful tools in helping to understand the customer base. However, these tools are often misused when retailers attach a segmentation code to a list of customers and then profile with the national average as the base. The national average is not an appropriate base for a brick-and-mortar retailer, since it makes sense that only people within a reasonable distance of a store are eligible to shop there. If a retailer can limit the base for the profile to only those eligible to shop at the stores, they’ll obtain a much more accurate customer profile.
9. Failing to recognize non-linear relationships
Every retailer knows that there is a one-to-one relationship between adding square footage to a store and increased sales, right? Not so. In fact, most retailers experience the opposite to be true. Despite that, countless models use a linear equation for square footage or other factors that are clearly not linear relationships. Using curve estimation techniques or non-linear regression can help to fit the model more accurately to the data.
10. Using inexperienced modelers
Some modelers say they can build a database of store information and create an adequate model using statistical techniques -- without the benefit of having training in spatial concepts, retailing concepts, catchment patterns or markets. While a modeler certainly doesn’t need to visit every store and market in the model database, he or she does need to know how to evaluate the retail landscape, site considerations and catchment patterns or, at the very least, review the model with someone with firsthand knowledge of these factors.
Building reliable and actionable site models is an entirely achievable goal, but it is one that requires care and diligence throughout the modeling process. “Too often, easily avoidable mistakes like these create obstacles for businesses and their success,” said Wolfe. “With the added value of location intelligence solutions, many companies are seeking to make their forecasting process more scientific and accurate. Armed with the knowledge of common site modeling mistakes, retailers will know what to watch out for, whether they are building their own model or hiring someone else.”