Assortment breadth against depth
In a store with fixed shelf space, adding a line removes facings from another. Breadth and depth trade directly against each other, and in small-format retail the constraint is tight enough that the trade-off is the central assortment question.
Data can inform it, within limits that are worth naming.
What does breadth buy?
Coverage of more needs and occasions, which raises the chance that a shopper finds what they came for. In a channel where trips are frequently single-item and specific, that matters more than it does in a large-format store where the shopper browses.
The cost is that each line has fewer facings, which increases out-of-stock risk and reduces visual presence.
What does depth buy?
Availability and visibility on the lines that sell most. Fewer facings on a fast seller means it empties between replenishments, and an empty facing sells nothing regardless of how good the assortment decision was.
Depth also simplifies operations: fewer lines to order, count and manage, which is a real benefit in a store with limited hours.
What can the data say?
Which lines are earning their facings, measured by sales per facing where facing data exists or by velocity per stocking store where it does not. And which stores show out-of-stock patterns on their fastest lines, which indicates insufficient depth.
The second is the more available and the more actionable. Persistent gaps on top sellers is a depth problem with a measurable cost.
What can the data not say?
Whether a shopper who could not find their preferred line bought a substitute or left. Lost trips are invisible in transaction data by definition, which means the cost of insufficient breadth is systematically understated.
That asymmetry is important. The cost of too little depth shows up in the data; the cost of too little breadth does not, so a purely data-driven assortment tends toward narrow.
How is the asymmetry addressed?
With substitution analysis where it is possible, and with judgment where it is not. If a line's absence is fully absorbed by another, breadth costs little; if the category's volume falls when it is absent, breadth is doing work.
The natural stockout analysis is the practical tool here and it answers the question directly for the lines that have been out of stock.
How does this differ by store?
Considerably. A store with a narrow catchment and a repeat customer base can run narrower than one serving passing trade with varied needs. Applying one assortment across an estate ignores that, and store clustering is what addresses it.
What is the practical process?
Rank lines by velocity, identify the tail, test substitution on the tail, and reallocate space from lines whose absence is absorbed to lines that run out. That sequence uses available data and it is repeatable.
Frequently asked questions
Is more choice always better for the shopper?
Not beyond a point, and the evidence on choice overload is mixed enough that it should not be asserted confidently. What is clearer operationally is that lines nobody buys occupy space that could serve lines they do.
How do I identify the tail?
By velocity per stocking store, ranked, looking for where the distribution flattens. The tail is usually longer than expected and it is where the space for depth comes from.
Does delisting a tail line lose the shopper?
Sometimes, and substitution analysis is what estimates how often. A line whose absence is absorbed loses nothing; one whose absence loses category volume was doing work that was invisible in its own sales.
How do I know if depth is insufficient?
Look for out-of-stock patterns on top sellers, particularly late in the day or late in the replenishment cycle. Consistent gaps at a predictable time are a depth or replenishment problem rather than a demand one.
Should assortment differ by store?
Where store behaviour differs, yes, and clustering stores by trip mix and category performance is the usual method. A single planogram across a diverse estate is a compromise everywhere.
What about new lines?
They compete for the same space, which is why a launch should be evaluated against what it displaces rather than against zero. That is a harder test and it is the correct one.
How often should assortment be reviewed?
To the buying cycle, and at minimum annually. Reviewing more often than the supply chain can act on produces analysis without consequence.
Does this apply to a wholesaler's range?
Directly, and the wholesaler's breadth decision constrains every store it serves. In independent retail that is frequently the more consequential decision.
What is the first analysis to run?
Velocity per stocking store, ranked, by category. It identifies the tail and the top sellers in one view, and both ends of that ranking drive the decision. Channel context is collected by NACS.
How does this change during a supply constraint?
Breadth becomes harder to hold and depth becomes the safer choice, because a narrow range reliably available beats a broad one with gaps. That is a temporary reordering rather than a permanent one and it should be reversed deliberately.
### What should a brand actually do with this?
Run the tail analysis and the out-of-stock analysis together, because they are two ends of the same decision. Space recovered from the tail is space available for depth on the lines that run out.
Present the trade explicitly to the retailer or wholesaler. A recommendation that frees space and names what it is for is considerably more persuasive than one that only asks for more facings.
Looking at the independent channel? See independent c-store insights and industry insights on nrsinsights.com.