By David Klein
Branch site selection is often thought of as a question of real estate. While real estate does play an important role, branch selection is first an analytical question for network planners. In other words: before you kick the dirt around at a proposed site, perform analysis to model the attractiveness of the location and how it impacts the entire branch network.
There are several approaches to modeling a proposed branch location. A screening model is a compare-and-contrast exercise. It allows you to compare a proposed site with the characteristics of your best branch to see how closely they match. A close match might mean a good site. However, a screening model doesn’t provide a sales estimate for the branch or assess the impact the new branch will have on the overall network.
Another approach, called a suitability model, adds demographic and market variables to the screening model to produce a sales estimate. The suitability model, like the screening model, does not account for the overall network or possible cannibalization of business from other branches. Using the screening or suitability model, you could end up choosing a new branch location that performs well, but at the expense of other branches, creating little net gain or even a loss in overall deposits across the branch network.
A network optimization model of bank branch selection not only helps you place new branches where total potential is high, it also balances new business at the proposed branch with cannibalization in the network. The model helps determine which proposed locations will increase overall deposits for the network – a primary goal of branch network expansion.
There is an important spatial – or geographic – component to a network optimization model. The model chooses locations based on the interaction between a customer’s distance from the branch (the spatial component) and the attractiveness of the branch.
Distance and attractiveness are the two main factors influencing a consumer’s banking decisions. Banking customers typically will patronize the closest, most convenient branch that can satisfy their needs. At some maximum distance, the customer will not choose your branch and will find another way to bank, most likely going with a closer competitor.
Branch attractiveness is a function of many attributes, such as the bank’s marketing, visibility, the presence of a drive-thru and even the management at a particular location. In a network optimization model, analysts quantify attractiveness and use it as an input into the model.
The first step in network optimization is to assess your bank’s current market position. This includes determining the bank’s overall market footprint, an area that encompasses all your customers. You also want to know your market penetration, which can be calculated by your share of households, often measured at the block group level.
The next step is to evaluate your bank’s competitive position. One way to do this is to license commercially available databases that estimate market penetration and potential for each savings, checking, investing, lending and insurance product. Estimates are based on consumer behavior reported in the Federal Reserve Bank’s Survey of Consumer Finances.
Next, overall market potential is analyzed. In a saturated market, opening another branch could cannibalize more of your own bank’s business than it takes away from competitors. On the other hand, untapped potential may indicate an opportunity to open multiple branches.
Now that you have a sense of your bank’s current market and competitive position, along with market potential, you can evaluate potential locations in the context of the overall market and network.
For any proposed location, you would create a trade area around it that mirrors trade areas around existing branches. For example, an existing branch may have a 10-minute drive time trade area that encompasses all of its customers. This same trade area calculation can be applied to a proposed location, and the demographics and financial market data within that trade area are analyzed to estimate deposit potential.
The key question to answer for any proposed branch location is how a new branch will fit into and impact the overall branch network. The network optimization model calculates the “patronage probability” and deposit value of the proposed site. Another way to express this is “people flow” and “dollar flow.” What is the likelihood that people will use the bank? What is the value of their deposits? During this phase of the analysis, the model can predict which existing branches will lose a percentage of business to the new branch.
It is natural that existing branches will lose some business when a new branch is added to the network. However, the new branch must more than make up for it – and if it does, overall deposits increase.
Once you know which areas are best within a market, you can put your real estate team to work identifying suitable properties in the area. Remember: first model the network, then pick the site.
How to Get It Done
There are basically two approaches to implementing a network optimization model for branch site selection. If you have a network planning department, you can license network modeling and site-selection software along with specific demographic and market data sets. Even with experienced analysts on staff, you may need training and support from your vendor to develop the right analytical model for your bank.
The other approach is to outsource network modeling to an expert partner, one with deep expertise analytics and proven success in helping banks select profitable new sites.
Either approach can work and either can deliver positive return when your new branches contribute to increased deposits and profitability of your branch network.
David Klein is marketing director for Rochester-based Mapping Analytics (www.mappinganalytics.com), a consulting services, software and data company that works with banks throughout the Northeast.