Improving sales per square foot in retail

The biggest department returns the least

Sales per square foot is the metric that decides whether a store's floor plan is earning its rent, and in most grocery formats it delivers the same unwelcome finding: center store occupies the largest footprint in the building and returns $288 a year per selling square foot against produce at $812.

That gap is not a scandal, because those departments are not interchangeable and center store does work that its own sales line does not capture. But it is the correct starting point for any conversation about space productivity, and most chains do not have the number to hand.

Comparing departments fairly

A raw sales-per-square-foot ranking will always put perishables on top and center store at the bottom, and acting on the raw ranking alone produces bad decisions. Three adjustments make the comparison honest.

Use gross margin dollars, not sales dollars. Produce carries shrink that center store does not. A department returning $812 in sales at a 32% margin rate after 8% shrink is returning less than the headline suggests, and center store's $288 at a 26% margin with negligible shrink closes some of the gap. Margin per square foot is the number that maps to profit; sales per square foot is the number that is easy to get.

Charge the right square footage. Selling square feet, not gross. A department's backroom and prep space is real cost, and departments differ enormously in how much of it they consume. Produce and bakery carry prep areas that center store does not, and excluding that space flatters exactly the departments that look best on the raw measure.

Account for the trip. Some departments drive the visit and others monetize it. If a department is the reason a basket exists, its own per-foot return understates its contribution. This is the honest defence of center store space, and it is testable rather than rhetorical: measure basket attachment and see whether the trips that include the department differ in total value.

Bakery is the finding. On sales per square foot it sits third and unremarkable. On margin per square foot, after its heavy shrink, it is close to produce and clearly ahead of dairy, which changes what you would do with an extra twenty feet of floor.

What actually moves sales per square foot

Sales per square foot has exactly two levers, and confusing them wastes resets.

Raise the numerator: sell more from the same space. This is assortment, pricing, availability, and merchandising work, and it is where nearly all the durable improvement comes from.

Shrink the denominator: use less space for the same sales. This is a floor-plan decision, it happens at remodel frequency rather than reset frequency, and it is where the largest single-step gains live.

Availability is the most under-used lever on this list. An out-of-stock top-decile item earns nothing per square foot while occupying exactly the same footage, and in most chains availability on the fastest movers is the cheapest sales-per-square-foot improvement available. It requires no reset and no capital, only the store-level inventory work to see where availability is actually failing.

The center-store question

Every chain eventually asks whether center store should be smaller, and the question is usually framed badly. "Center store returns $288 and produce returns $812, so convert center store to produce" ignores that produce demand is finite and that center store carries the shopping mission that brings people in for the perimeter.

The better framing is marginal rather than average. Not "which department returns more per foot," but "what would the last twenty feet of center store return if it were something else, and what does it return now." Average returns are a bad guide to a marginal decision, and the last twenty feet of center store is almost certainly returning far less than the department average, because that space holds the slowest tail of the range.

That reframing also connects to the overstock finding: the slow center-store tail carries 6.3 weeks of supply against a 3-week ceiling. Space occupied by inventory that turns four times a year is the specific space worth converting, and it is identifiable rather than theoretical.

Measuring it without fooling yourself

Two measurement traps produce sales-per-square-foot numbers that cannot be compared to anything.

Seasonality against a fixed denominator. Sales move seasonally and square footage does not, so a quarterly sales-per-square-foot figure swings with the calendar. Compare against the same period last year, or annualize, but never compare Q4 to Q2 and conclude anything about productivity.

Format mixing. A 12-store small-format cluster and 18 full-format stores have structurally different returns per foot, because fixed space (checkout, aisles, back of house) is a much larger share of a small store. Chain averages across formats describe nothing. Compute within format, and compare a store to its format peers.

Both traps produce the same symptom: a number that moves for reasons unrelated to any decision anyone made, which teaches everyone to ignore it.

Why the metric gets ignored inside stores

Sales per square foot is a headquarters metric, and store teams almost never act on it, for a reason worth taking seriously: at store level it is not actionable. A store manager cannot change the footprint of a department. Presenting them with a departmental space-productivity number is presenting them with a fact about a decision made in a remodel three years ago.

What is actionable at store level is the numerator, expressed in terms the store controls: availability on the top items, compliance with the planogram, condition and rotation in the perishable departments, and speed of restocking after delivery. Those roll up into sales per square foot and each one is a thing a team can do on a Tuesday.

The practical consequence is to keep the metric at the tier where the decision lives. It belongs on the board and executive tiers, where floor-plan and capital decisions actually get made, and its store-level translation should be the operational drivers rather than the ratio itself. Sending the ratio down the organization produces the familiar situation where everyone can recite the number and nobody has ever changed anything because of it.

Benchmarking against yourself first

External sales-per-square-foot benchmarks are widely quoted and mostly unusable, because format, region, department mix, and whether the published figure uses gross or selling area all vary and are rarely stated. A chain comparing its $288 center-store figure to a published industry number is usually comparing two different measurements.

Internal benchmarking is both more available and more useful. Within a format cluster, the spread between the best and worst store on the same department is a direct measure of how much is achievable without any capital, because those stores have the same footprint and range. If the top store in the 18-store full-format high-volume cluster returns materially more per foot in dairy than the median, that difference is operational, and it is reachable.

That spread is also the right basis for a target. A goal derived from the 75th percentile of your own format peers is defensible, achievable, and immune to the definitional problems that make external comparisons an argument rather than a measurement.

Turning the number into a decision

The gap between measuring space productivity and acting on it is where most of this work stalls, because the actions live on two very different clocks and get discussed together.

Separate them explicitly. The reset-cycle actions, tail cuts, category reallocation, and availability work, should be running continuously and need no approval beyond the category team. They are where the compounding improvement comes from, and treating them as part of a space-productivity initiative slows them down by attaching them to a capital conversation.

The remodel-cycle actions, footage reduction and space conversion, need a business case, and the business case needs the marginal number rather than the department average. Building that case is a discrete piece of work that happens once per remodel cycle per store, and it should draw on the last several years of category-level performance rather than a snapshot.

Chains that mix these end up doing neither well: the operational improvements wait on a capital decision that takes a year, and the capital decision gets made on averages because nobody prepared the marginal analysis.

Doing this in Scout

The reason most chains cannot act on this metric is that the space side of it lives outside every analytics system. Sales are easy; square footage sits in a fixture or planogram system, sometimes in a spreadsheet, often only as gross store area rather than by department.

Scout carries the department and category space allocation alongside sales, margin, and shrink, so margin per selling square foot is computable rather than estimated. Comparisons are scoped to format cluster by default, which removes the mixing trap, and the seasonal comparison is year-over-year rather than sequential.

The more useful output is the marginal view rather than the average. Ranking space by the return of the items that occupy it identifies the specific low-return footage, which is what a remodel conversation actually needs. A department average tells you center store is weak; the marginal view tells you which twenty feet.

Scout measures space productivity and ranks the opportunity. It is not a space-planning, fixture, or store-design system, and it does not produce floor plans.

It is also worth stating the denominator on the face of every report. Whether a figure uses selling or gross area changes it by a wide margin, and a chart whose basis is undocumented will eventually be compared against one computed the other way, which produces a conclusion that is confidently wrong.

Summary

• The largest department usually returns the least per foot. Center store at $288 against produce at $812 is the normal starting point, not a scandal.
• Adjust for margin, shrink, and selling versus gross footage before ranking. Doing so moved bakery from third to effectively second and changed the answer.
• Availability on top-decile items is the cheapest lever, and marginal return, not department average, is the right basis for any space-conversion decision.

Further reading: shelf space optimization covers reallocation within a department, and assortment optimization covers the tail that occupies the low-return footage.