THE PREVIOUS TWO CHAPTERS delved deeply into the methodology a retailer should follow to expand into new markets. This chapter provides specific examples of work that we have done with businesses to find the best metropolitan areas for expansion and the best locations within those areas. The actual data is disguised for reasons of confidentiality and has been consolidated to portray a single fictitious company, but the problems are typical of many retailers and the analysis tracks exactly the process we have followed many, many times. The numbers used to describe Metropolitan Statistical Areas (MSAs), the number of stores in various trade areas, the number of competitors, and similar data are purposely chosen to represent ranges that apply to many companies. The same is true for the dollar amounts used in store sales. Obviously, numbers for your concept could vary considerably, but any demographic analysis you commission should achieve the same depth of analysis, obtain a similar level of detail, and enable a well-informed set of decisions about expansion and relocation.
Our featured company is Opportunity, Inc., which sells high-end consumer soft goods. It has a number of stores in both small and large markets, and the performance of those stores has been spotty. Like many other retailers, its expansion had been somewhat helter-skelter. Before embarking on a major expansion, the company sought a complete reexamination of its real estate selection process. “Save us from ourselves” is how one company executive explained the task.
It is not sufficient to rank potential expansion areas in order by population size or income levels, although both elements play a role and income is often a deciding factor. The goal is to create identifiable tiers according to your selection criteria so that you end up with a reasonable number of top MSAs on which you can initially concentrate. Some analysts create two tiers. Our own methodology is to use four. We find that a model with four tiers does a good job of predicting store performance, particularly in tiers 1 and 2. The specific cut point for income to create a tier depends on the concept. A concept appealing to a lower income demographic would use a lower income number to sort the tiers; a concept appealing to a higher income demographic would use a higher income number.
Because the demographic is defined by other factors before tiers are created by income, sometimes the cut point varies considerably even for concepts that appeal to customers with similar income levels. Choosing the cut point therefore requires great care. Tiers defined by the number of customers with $100,000 in annual income, for example, would yield many fewer tier 1 markets than tiers defined in terms of $50,000 in annual income. The smaller yield might be positive if you needed to create focus, or if the product itself was expensive, such as with luxury vehicles or yachts. The smaller yield would be negative if the goal was rapid expansion because the model would generate too few target markets to fuel high growth. A high cut point might also create too many lower quality markets in tier 2; that is, you could end up with a tier 1 of ten markets and a tier 2 of 50 markets, which is too many to effectively prioritize. The same thing can happen if the cut point is too low—so many markets appear in tier 1 as to make the division meaningless. It can take time and effort by a skilled analyst to find the right balance; for example, the right balance might be a model that yields ten tier 1 markets and 20 tier 2 markets.
Our analysis for Opportunity, Inc., quickly identified four discrete market types. Tier 1 was an MSA with greater than 400,000 people who had high income, which equates to a propensity to spend. (Remember, the 400,000 figure is not an arbitrarily large number, but a cut point that comes directly from the demographics for this one concept and the need to create manageable tiers.) Tier 2 was an MSA with greater than 400,000 people and lower average income. Though tier 2 had fewer propensities to spend, the sheer size of the population provided enough concentration of high-income earners to make the areas attractive. Tier 3 was an MSA with high income and low population. Tier 4 was low population and lower income. The top two tiers provided 80 percent of Opportunity's revenue, and this was not because the top two tiers had more stores. In fact, Opportunity had only 67 stores in the top 75 MSAs.
Further analysis showed no regional differences in sales, but a wide variation according to tier. Nearly two-thirds of the weakest stores (which accounted for nearly half the total number of stores) were located in tier 3 and 4 MSAs. The weakest stores earned barely half the chain average. The per-store revenue for tier 1 and 2 stores was 20 percent higher than the overall chain average and was nearly three times that of the weakest stores. Further, we found a strong correlation between store revenue and the number of high-income households within eight miles of the store. More than one-third of Opportunity's stores, however, were located in areas with low income. The implication was that new locations could dramatically improve sales for a number of stores.
Our analysis showed how competitive their particular category is. Of the top 25 markets, five had 100 or more competitors. Another ten MSAs had 50 or more. Most others had between 25 and 50 competitors. Only a few MSAs had a small number of competitors, and these were not close to where Opportunity was well established and could support operationally. These latter MSAs would provide Opportunity a chance to capture untapped markets in future years, but they would not enable the company to improve its market position quickly. Overall, competitors left few openings to be exploited. Falling back on the fundamental idea of “fishing where the fish are,” we recommended that Opportunity's short-term emphasis be to own a particular state—in this case, Texas. Despite the presence of strong competitors, Opportunity's most profitable stores were there and it had the operational capability to support more. Four of its tier 1 MSAs and five of its tier 2 MSAs were located in that state. Its strategic drawback (here and elsewhere) was not the strength of its competition but an insufficient concentration of its own stores.
“Owning Texas” and bolstering its presence in the other top MSAs in the country would maximize sales as well as operational efficiency by putting a number of stores in close proximity. By siting 50 new stores in three years in MSAs with the greatest potential, we estimated, Opportunity could improve its sales by 13 percent per year, even if the new stores did no better than the current average store sales. If Opportunity could double the number of stores in three years in the correct MSAs—a stretch, but it was possible—and saw modest revenue increases in the new stores as the result of better locations, the company could increase its overall sales by at least 24 percent annually.
Prioritizing MSAs according to customer profiles and operational strength gave Opportunity an orderly way to proceed over the next several years as well as the greatest opportunity to turn around its retail business. Opening new stores provided the twin opportunities of expanding the core business while upgrading the brand through new store designs. Using the evaluation criteria developed for new stores, Opportunity was also able to create a systematic method to determine the future of existing stores. Opportunity agreed to close poor performing stores in areas that had poor demographics and therefore little likelihood of future growth. If a store was in a poor location but in an area with good demographics, we recommended relocation at the time of lease renewal—unless relocation would generate more revenue than the cost of terminating a lease, in which case an immediate move would occur.
Because Opportunity already had a number of stores in smaller markets, you might wonder why we did not recommend that Opportunity target the small-town market as Wal-Mart did with general retail merchandise. (Wal-Mart followed the See Ya strategy in overwhelming existing small merchants.) The answer is that Opportunity sold to a broad range of customers. Its stores were operating in a representative sample of cities of all sizes. When we analyzed the customer base, we found that the customers who chose Opportunity over its competitors happened to be in major markets. In fact, Opportunity did relatively better in larger cities than in small. Thus, we recommended that Opportunity put new stores in top metropolitan areas. If the customers had been in Small Town, U.S.A.—or if Opportunity had determined to become the destination store in its category for small markets—we would have gone in that direction.
Figure 14-1. Based on the customer profile and competitive considerations for the company Opportunity, Inc., the city of Houston, Texas turned out to be a prime candidate for expansion. Demographic analysis identified a number of “hot spots” (darkest regions), indicating potentially good locations in the metropolitan area. Additional analysis would identify the best two or three trade areas for Opportunity's concept; Opportunity would select the best locations available within those areas. If this were the first store for the company in Houston, the specific trade area might not matter. However, if Opportunity already had stores in the city, then “hot spot” analysis would drill down perhaps as far as individual street intersections to determine the best places for the company to in-fill with stores without reducing sales at existing stores.
After we had chosen Texas as the targeted expansion territory, we examined each of the MSAs in the state. We identified Houston as the top market to enter based on the company's competitive and operational strengths. Using the methodology and demographics detailed in the preceding chapter, we developed a “hot spot” map of trade areas specific to that concept. The maps guided us to specific trade areas and sub-markets that contained good prospect sites based on the criteria we knew were correlated with store performance. After we identified these prospects, we compared them to existing analogue stores that had similar characteristics to validate our “on-the-ground” evaluation and to develop comparables-based sales forecasts. Finally, we were ready to examine the actual prospect sites to make a store selection. Chapter 3, “The Importance of the First Store,” discusses site issues specific to a street location. The following example discusses site issues specific to a shopping center.
Our analysis pinpointed two possible shopping centers as worthy of pursuit. We were charged with the on-site inspection and recommendation as to which location was preferable. The first shopping center had good visibility and decent traffic counts, but it lacked national brand co-tenants that created high-frequency shopping that best supported Opportunity's concept. In addition, the shopping center was in the middle of a block, so it had limited turn access into it. The second shopping center had higher traffic counts and several national co-tenants that were attractive for the concept. In addition, the second center was located at an intersection with a signal, making access easy. These factors enabled us to quickly eliminate the first shopping center from consideration.
Another advantage of the second shopping center is that it offered three separate possible locations for the Opportunity store, so we visited the site to see which of the three potential areas of the shopping center would be best. All of the issues with the three sites and with the shopping center itself are representative of the issues that any retailer is likely to find in a typical shopping center, so we'll examine the possibilities in detail.
Figure 14-2 shows the three retail spaces available to rent at the center in the locations marked A, B, and C. The question for the client was which of the three would be best for a general retailer. In order, we looked at traffic, access, and co-tenancy. From our traffic database, we found that Rocket Road, running north to south on the left side, is a major artery carrying 50,000 vehicles a day. Yao Parkway, running east to west across the top, carries about 30,000 vehicles a day. The map of the shopping center had already showed us several things. Site A was in a corner location on the main building, close to Lum's Restaurant and Walgreens Drugs, which drew high-frequency customers, but there was no entrance close by. Site B had good access and visibility and had high-traffic stores on either side, but the store shape was long and skinny. Site C was located on Rocket Road, by far the busiest street, and had great visibility, but access was limited from Rocket Road. Touting its features, the shopping center landlord was eager to rent Opportunity one of the spaces at Site C. This space rented for $30 per square foot versus $24 for B and C, and usually rent is proportional to the quality of the space.
Figure 14-2. After a potential location is identified, a physical inspection is required. The map indicates that Site C would be the best location for Opportunity, Inc., because of the proximity of the two major cross streets. However, an on-site visit made it clear that Site B was superior to both Sites C and A. Among other factors, the visit revealed that Site C was ten feet below the grade of Rocket Road and hard to see.
Many things jumped out at us during a personal inspection of the site that were not evident from the map. For Site A, we saw that Lum's and Walgreens took up all of the limited parking on the east side of the shopping center. Also, the Walgreens building would block Site A's store signage from Yao Parkway. For Site C, we saw that a lot of the traffic on Rocket Road came from people traveling east on Yao Parkway and turning north onto Rocket before they reached the center. The 50,000 vehicle count dropped dramatically on Rocket Road south of the intersection by the shopping center. The on-site inspection also showed why there was no access to the site from Rocket. The retail building was ten feet below the grade level of the road. Not only would customers not have access from the road, but a store also would not be nearly as visible from Rocket as we expected. (When a map shows the lack of an entrance on a street, there's always a physical reason, such as grade, or a practical reason, such as traffic congestion.)
For Site B, the only drawback was the shape. Office Depot and Trader Joe's, two high-frequency co-tenants, were on either side. The site was only a 100 feet from Circuit City, which draws customers from the overall region. From the map we were leaning toward Site C—and the landlords were guiding us that way—but an onsite inspection made it clear that Site B was really the best for Opportunity's needs. You can work around most space problems, but you can't get around the visibility and access problems. Best of all, Site B was $6 a square foot less!
Thisexample shows why a personal inspection of the site is mandatory. Every location has some kind of physical issue, good or bad, that cannot be discerned from maps, schematics, or aerial photos.
One final issue. What if Site B were not available? Would we recommend that Opportunity take Site A or C? What if no sites were available at all in this shopping center—would we recommend the less desirable site at the first shopping center? There's no perfect answer. Because we have already done the analysis, we know that all the sites should be “good enough,” capable of yielding a profit, so the expectation is that the better site will yield more profit, not that the lesser sites will result in a loss. This assumes, of course, that our site inspection did not determine that the physical shortcomings of one of the other sites did not render our projections meaningless. Let's say the location you want could generate a 15 percent return but will not be available for three years. Let's say your on-site inspection causes you to drop the estimate of the return on the secondary location to ten percent, but the revenue begins this year. There is value in having less money today rather than more money in the future. Too, you have no way of knowing whether you will be able to obtain the better space in three years anyway, or at a price that will make it more profitable than the second choice. (If the prime location becomes available, then your problem is whether you need to take the space for strategic reasons, to prevent a competitor from obtaining it.)
If, after seeing the sites first hand, you still have confidence in the financial projections for each location, then the next consideration is whether you absolutely, positively have to be in this trade area. In most cities, two or three trade areas provide the most brand visibility. They are so strong and have so much trade activity that you simply have to be there to get brand exposure and to succeed financially. This is the one exception in which a small retailer might take a B location in an A trade area. Another issue is the strength of the particular shopping center or mall. In Seattle, Washington, University Village is so powerful that you would think long and hard before saying no to a space, even if it were not the one you wanted. In San Ramon, Calif., one shopping center is heads and shoulders above all the others in the trade area. You would take a small space at the preferred center, even at less than favorable terms, over any other spaces, on any other terms, at the other shopping centers. It's that, or find another trade area.
Objective guidelines such as these make it easier to balance the ever-present trade-offs and provide you a greater comfort factor in making difficult site selections.
Because Site B was the preferred choice for a high-concept consumer brand, would other concepts do as well in Sites A and C and at the other center on Rocket and Yao? The answer is yes—a retailer with an appointment-oriented business. Site C has a nail salon and a company that makes window blinds. An optometrist would be a logical addition, or an insurance agent. Any of the shopping center sites would work for a business in which a customer would most likely search it out rather than drop in on a convenience trip. Site C also has a nutrition center, the kind of business that customers are normally willing to seek out.
Similarly, any of these sites will work for a local destination, such as a sandwich shop, where customers will come back regularly once they discover you. (Site C has a pizza shop.) The sites will also work for a regional destination, where people will drive some distance for your products. All of western Washington State has only one IKEA furniture outlet, and people will find it. Wal-Mart draws people from an entire county. Nebraska Furniture Warehouse in Omaha draws people from the entire state. Outlet malls are probably the best example of a regional destination that can be situated almost anywhere. They locate along freeways between towns or on main highways in out-of-the-way towns to avoid competing directly with mainline stores selling the same national brands, but customers will drive from Portland or Seattle to, say, the Oregon coast or the Washington interior to seek out bargains.
As the previous examples show, you can get away with a secondary site at a good location—if you meet a need that the customers alreadyknow they have . Most retailers, however, are convenience-oriented locations. People will drive ten miles to go to The Home Depot, but they will drive only one mile to go to a dry cleaner. People won't drive out of the way to buy milk or gas or most other commodities. If you sell specialty items and you're off the beaten path, how will customers even know you're there?
A few retail categories may fall in between convenience and destination. Starbucks began as a convenience location—it had to be “on the way”—but as the company became ingrained in American culture many Starbucks stores turned into local destinations, places that people seek out in their neighborhood for a break in the day. Radio Shack is a cross between convenience and destination. Some traffic consists of convenience drop-ins who want to check out the latest electronic toys and other gifts, whereas other traffic consists of destination customers who come with a specific mission—to get a connector or battery or similar part.
Knowing exactly who your concept appeals to and where it fits into the destination or convenience category can be difficult for some new enterprises. The problem is compounded by what we've seen with every location having some kind of “gotcha” or some kind of “aha!” Stuart Skorman, the founder of Elephant Pharmacy, the high-service, high-touch drugstore in Berkeley, California, found out within a few months of opening that he needed more local drawing power. Although the target market was perfect demographically—a well-educated and upscale Berkeley neighborhood—and the store site was physically perfect within that demographic base, the community simply wasn't large enough. There were just not enough people in the one- to three-mile range to provide the business that Elephant needed, based on the draw of a traditional pharmacy. Stuart had selected the site because it conformed to the Walgreens model of “local” site selection. Like Starbucks, Walgreens saturates a market with a number of local stores. As it turned out, his great local base of customers protected his downside but did not maximize his upside. A supermarket offers many reasons for customers to come once or twice a week, but the core of a pharmacy's business, prescriptions, is a monthly draw. Because it can take as long as five years to change a customer's prescription-buying habits, a new pharmacy brand will not necessarily see a rush of business upon opening. Elephant therefore had to work hard to draw in customers more often. Although the company needed more consumables to drive more visits, the pharmacy also had to stay true to its core demographics. Where a Walgreens' food mart offers frozen foods, the town of Berkeley doesn't “do” TV dinners and beer. It does fresh fruit and organic vegetables. Elephant had to choose consumables to match the location—the neighborhood—in order to entice people to come daily. In addition, Stuart lowered his margins to increase traffic and started a direct-mail catalog touting all of the free classes, which sparked a surge in interest. The first four or five months were hard, but the store has slowly generated higher traffic. That Elephant Pharmacy did not map to a strict definition of “local,” “convenience,” or “destination” proves that every entrepreneur needs to carefully study the initial reaction to a concept. Retailers should be prepared to rapidly adapt the store's offering to the actual needs of the customers in a particular location.
A sophisticated demographic model can drive intelligent decisions regarding expansion, and on-site inspections of potential locations help uncover any street-level issues with a site. The final piece, then, is to be sure that the site maps to the kind of destination that your concept actually is. A location problem, as Stuart says, is an expensive lesson to learn. Once again, we see that location will make you or break you. For all the possibilities of being local or regional destinations, though, most retailers are strictly convenience retailers. There's no real issue about distance or the need to become a destination. For 90 percent of retailers, customers have to see your store—before they see your competitor's store—and be able to get to your store easily. Visibility, strategic location, and access are everything. You have to make it easy for customers to find you and stop because most retailers meet a need that customers don't know they have until they see your store.
As these last three chapters have shown, consistently locating and securing good real estate for a retail concept is exhaustive and time-consuming work. But there is more to it than segmenting the market, analyzing the demographics and finances, and pounding the pavement. Real estate acquisition is a process, but real estate strategy is an art form. Whenever I prepare to talk to someone about real estate expansion, I show them the Jasper Johns painting “The Map,” which I have as a reproduction on the wall in my office entry. The painting consists of a large, irregularly shaped object worked through with thickly daubed primary colors. The colors are so brilliant and the texture is so interesting that you see the object as a whole, as an abstract painting that seems strangely familiar. It requires a careful look to recognize that the painting is actually a map of the United States, with each state boldly colored and clearly labeled. A typical U.S. map for a retail chain, which shows stores as dots in different cities, is informative but not beautiful. Johns' painting is art, and it has become an icon for a thoughtful real estate strategy. You don't look at real estate as separate pieces. Rather, you visualize the end result that you seek, in completed form and in terms of the pieces that must fit together to roll out a retail concept. People can identify with the beauty of the painting and begin to understand that what we're doing is more than just a job or the drudgery of process. Real estate strategy requires that you always keep in your head the totality of what you are trying to create and that you think of the process as a series of interlocking events. The steps lead to particular locations, so it is easy to think about them separately. The painting is a visual reminder that the mission of real estate is to pick all the right pieces so that they can be quilted together to create a single, distinguishable—and distinguished—brand.