Marginal Return on Space – an Optimization Primer
If you have been in Space Management for any amount of time, you are familiar with GMROS (Gross Margin Return on Space.) Calculated roughly as the (gross margin $ ÷ floor space) of the product hierarchy being measured. What that gives you is a stale snapshot of the total dollar contribution of the department, category or planogram. But if you dig into that metric you will learn that there are sections of the department or planogram with much higher GMROS than the department or planogram average. Which brings us to marginal return analysis in space optimization.
A question that I hear from the retail C-suite is “How can I succeed in a small store footprint?” Real Estate wants to know that the optimal minimum prototype size for leasing purposes. Store Planning wants to know how to shoehorn big store sales into small store footprints. Concept Development wants to know how much space is needed to support baseline sales and how much they can devote to experience. The key is Marginal Return on Space.
Let’s imagine you have a chain of 9,000 square foot stores. To serve inner urban locations and afford the rent in cities like New York and Boston, you need to develop 4,000 square foot stores. Going to your merchants, you are likely to develop a set of stores with one facing each of the top few hundred SKU’s in an incoherent merchandising mix. There is something I call the Point of Customer Relevancy that you need to reach with each category to have customers give you credit for being in the business. For example, you could create a 4′ household paper section with one SKU each of paper towels, napkins, toilet paper, paper plates, sandwich bags, aluminum foil, etc. But because you have not reached a Point of Customer Relevancy in the household paper category, customers wont hold your new stores top of mind as a place to purchase those needs. You will become a convenience purchase on par with gas station c-stores.
Still, I see retailers make this decision over and over again when they create small stores. No one wants to eliminate business lines and create stores that don’t offer the full range of assortment from their big stores. So what they get are small stores with a smattering of every product line that fall short of customer expectations and revenue projections.
From a purely analytic point of view, retailers can build a space optimization model that creates stores from the ground up using Marginal Return on Space. To begin, parse out GMROS metrics to a level more granular than department or category. In shelved goods, you can use the GMROS metric for each 4′ section of a planogram. In other areas, use the GMROS metric for each apparel fixture, produce or meat segment or glass counter.
Then, build the model to begin by saying “if I only had a store that was 12 square feet (4’X 3′ – the standard space allocation for a planogram section including shopping space), what would it be?” The model will select the most productive segment among all available. Then subtract that productivity from the total and ask the same question “what if I add another 12 square feet?” The model will build out the store to create the most productive store possible and you can stop the model at any point.
When I did this years ago for Best Buy, we quickly realized that if you only had a 1000 square foot store, it should be nothing but cell phones and accessories. (Mall kiosks, anyone?) Our store had to be gargantuan to justify any space at all for appliances or classical music. What ultimately became Best Buy’s Mobile stores were predicted by this model. As long as the marketing team works with the store development team to ensure that customer expectations about assortment can be managed, the model builds an extremely successful model. (Which is why Best Buy created the monicker Best Buy Mobile.)
This model will create optimal space allocation. It will still require an experienced hand to take the space allocations and create the best location layout and adjacencies. (These are the three sisters of macro space optimization.)
So if it’s so easy, why don’t more retailers use this tool?
- A lack of knowledge about macro space optimization and how to apply it to real world decisions. Macro Space geeks are rarely in positions that allow them to have a voice at the table when strategic decisions about prototype development and store strategy are being made.
- A lack of metrics in non-planogrammed areas. Apparel, seasonal and fresh foods are often high margin areas with little to no planograms. Retailers who rely solely on standard software tools usually become frustrated while developing metrics for non-planogrammed areas. Without those metrics (at the fixture level) the model doesn’t work. (for more info see: How to Make Macro Space Work.)
- A need to appease every business line. Despite data that proves that customers are not shopping all departments in equal numbers, retail executives buy into the notion that every business line needs “fair representation” even in small stores. Whether forced by internal politics, vendor marketing funds or promotional plan restrictions, most retailers will make the easy (and less effective) choice to include a single SKU from every business line rather than eliminate them from small stores.
Once this model is created it can be used for more than creating new small store footprints. Based on real estate pro forma’s, optimal store sizes can be determined to maximize investment payback. Gross Margin can be replaced with future business plan forecasts to model what the store of the future should look like. It can be used with store-level data to anchor remodel plans in an optimization methodology.