By Bill Scott – StoreReport LLC
In a chain of convenience stores consisting of 10- 100 locations, within a geographical area covering 50- 150 miles, their suppliers assume all stores within that area have similar clientele, when in actuality each store, according to their individual demographics is unique, and many of these stores are dying because they are offering the wrong product mix to a varying clientele.
Suppliers and distributors (and often a larger chain’s own central merchandising team) frequently treat a 10–100 store group as essentially homogeneous. They push a largely standardized “regional” or “cluster” assortment based on average velocity, broad demographic proxies, or the easiest logistics model. In reality, even stores a few miles apart can serve distinctly different clientele—different income levels, ethnic mixes, age profiles, work patterns, traffic types (commuter vs. residential vs. industrial vs. tourist), and daypart needs. The result is exactly what you’d expect under these circumstances: the wrong product mix in too many locations, soft sales, excess slow movers, out-of-stocks on the items that actually matter locally, and stores that gradually decline. Learn more about this in Small Business Retails’ Last Chance.
Why Suppliers Default to This Approach
• Operational simplicity and cost. Managing true store-level variation across even a modest chain multiplies the work of planograms, ordering, inventory allocation, and category resets. o Pulling off old or discontinued products o Adding new items o Changing the number of facings – how many numbers of each product are visible o Rearranging the order and positioning of products o Adjusting shelf heights, or secondary displays if needed o Updating price tags and signage.
A single or small number of “regional” sets is far easier for the distributor’s systems and salesforce.
• Data limitations on their side. Distributors often work primarily from their own shipment data rather than the retailer’s full POS + demographic data. Shipment averages smooth out local differences. • Incentive structure. Category captains and sales teams are rewarded for volume and distribution of their brands across as many doors as possible, not for perfect local fit at every door. • Legacy habits. The industry has long operated on route-based or regional models. True hyper-localization has only become practical with better data tools in recent years.
Even when suppliers offer “clustered” assortments, the clusters are often too coarse (e.g., “urban” vs. “suburban” or simple size bands) and fail to capture the real differences you see store by store.
The Damage
When the mix doesn’t match the actual customers:
• High-velocity local demand items are under-spaced or missing → lost sales and customer frustration. • Irrelevant or low-velocity items occupy valuable facings → capital tied up, higher waste, and weaker margins. • The store loses relevance in its neighborhood → declining traffic and a slow death spiral that is hard to reverse once competitors or alternative channels capture the customers.
This is especially painful for chains in the 10- 100 store range: large enough that pure independent flexibility is gone, but small enough that they rarely have the sophisticated category management infrastructure of a national player.
What Actually Works to Fix It
Core Principles
• Customer demand and sales data first: Decisions start with what actually sells. Retailers analyze POS (point-of-sale) data for velocity (units sold per week), sales rankings, stockouts, and customer requests. Core staples in snacks, beverages, tobacco/nicotine, candy, and foodservice earn permanent space only if they perform consistently. Slow movers get cut. Top performers (often the top ~20% of SKUs driving most sales) get priority facings and never go out of stock. • Local demographics and trip missions: Assortment is tailored to the neighborhood and shopper types (commuters, night-shift workers, families, truckers, students). A store near a campus differs from one by a highway or in a residential area. Daypart analysis matters—morning coffee + breakfast items, afternoon snacks/energy drinks, evening prepared foods. Localization tools and AI help larger chains refine this by store. • Space productivity and limited SKUs: C-stores typically carry far fewer items than grocery (often 2,500–3,500 SKUs). Note: I have worked with operators that typically carry from 4K to 7K unique items of inventory in a space of less than 3,000 square feet. Every inch of shelf or cooler space must justify itself via sales velocity, gross margin dollars per linear foot, and inventory turns (often targeting high turns of 12–20+ annually for many categories). Over-assortment creates waste and slows movement; retailers balance core items with controlled variety.
Key Evaluation Criteria for Products
Retailers (and their distributors/category managers) weigh:
• Sales velocity and turns — Does it sell fast enough? • Profitability — Gross margin percentage and, more importantly, margin dollars generated relative to space and inventory cost. High- margin impulse items (candy, certain snacks, accessories) and foodservice often outperform low-margin traffic drivers like tobacco. • Category role — Destination items (e.g., key beverages or prepared foods) get broader selection; impulse or complementary items get targeted placement. • Shelf life and waste risk — Prefer items with good longevity or high turnover to minimize spoilage, especially for fresh/prepared foods. • Fit with basket and cross-merchandising — Products frequently bought together (coffee + pastry, chips + soda, prepared food + drink + snack) get prioritized and placed near each other. • Trends, seasonality, and innovation — New or trending items (better-for-you options, energy drinks, oral nicotine, local products, seasonal goods) are tested if they can drive incremental sales or attract new customers. Vendor pitches, syndicated data, trade shows, and proprietary analytics inform this. • Supplier terms and logistics — Pricing, margins, delivery reliability, return policies, and support from distributors (common in c-store wholesale) influence decisions. • Competitive and experiential factors — Gaps vs. nearby stores, ability to “surprise and delight,” private-label opportunities, and alignment with promotions or loyalty programs.
Typical Process
1. Review historical sales, category performance, and market data (including fair-share analyses of brands/subcategories). 2. Define category roles and target assortment depth/breadth. 3. Rank items and decide adds/deletes—core best-sellers stay; underperformers exit; promising new items trial. 4. Allocate space via planograms (shelf layouts optimized for velocity and visibility—eye- level for top sellers, impulse zones at checkout/endcaps). 5. Localize by store where possible and monitor continuously (weekly/period reviews). Adjust for seasons, trends, or performance shifts. 6. Use tools ranging from basic POS reports to advanced analytics/AI platforms for assortment optimization and localization. (More about this in coming articles.)
Distributors and category managers often play a major supporting role, providing recommendations, data, and planograms, while independent operators have more flexibility for local or novel items. Larger chains increasingly use technology for predictive, customer-centric decisions.
In short, the goal is not maximum variety but the highest-performing mix that matches quick-trip shopper needs, maximizes profit per square foot, and keeps the store relevant and in- stock on the items people actually come for.
Progressive operators (and the better distributor partners) are moving toward genuine localization:
• Store clustering based on real data, not just geography or size. Effective clusters combine POS velocity patterns, census/geodemographic data for the trade area, traffic type, and even competitive set. Two stores in the same zip code can belong to different clusters.
• In the average convenience store 30% of the items produce ALL of the profits, 55% to
60% of the items are marginal at best, and 10%-15% is dead and should be immediately removed from the store.
• Why do stores carry twice the inventory needed to meet customer service level?
1. Fear of losing sales on the core items – Retailers (and their suppliers) worry that if they cut too aggressively, they will stock out of a winner or fail to have the exact variant a regular customer wants. This leads to over-facing and over-stocking the good items “just in case,” plus keeping marginal alternatives as insurance. The result is excess inventory even while true service levels on the top performers are not always better, an
increase in theft and problems keeping the inventory organized, the shelves cleaned and the facing pointing toward the customers.
2. Supplier and Distributor Pressures
• Manufacturers constantly push new items and line extensions. Getting distribution is a key performance metric for their sales teams. • Distributors have historically been accommodating—stocking slow specialty items requested by individual stores or accepting manufacturer pushes rather than strictly enforcing velocity standards. • Slotting fees, promotional funding, free goods, and display allowances create short- term incentives to accept extra SKUs even when the long-term economics are poor.
3. The illusion that “more variety = better customer service” Many operators believe customers expect broad choice. Convenience shoppers prioritize speed and finding what they came for quickly. Too many near-duplicates often create decision friction and make the store feel cluttered rather than helpful. The extra SKUs rarely generate enough incremental sales to justify their cost.
4. Difficulty making hard decisions and lack of disciplined process
Cutting items requires saying no to suppliers, risking occasional customer complaints, and having confidence in the data. Many operators lack regular, rigorous assortment reviews based on true profit contribution (velocity × margin − carrying costs − space opportunity cost). Without that discipline, the assortment slowly bloats.
5. Operational and systemic factors
• Minimum order quantities and delivery patterns can force extra inventory. • Poor visibility into true store-level performance (especially when relying mainly on distributor shipment data rather than full POS + demographic analysis). • One-size-fits-all regional assortments that don’t match individual store demand (as noted earlier), so every store ends up carrying items that only work in a few locations. • Habit and inertia: “We’ve always carried it” or “The last manager ordered it.”
6. Misplaced Focus on Sales rather than true profit and capital efficiency
Many decisions still emphasize gross sales or category sales growth over contribution to store-level profit, inventory turns and return on space/capital. Slow movers can look acceptable on a simple sales report while quietly destroying cash flow and crowding out better performers.
The Net Effect
Stores end up with far more SKUs than needed for high in- stock rates on the profitable core. The extra inventory ties up working capital, increases shrink and waste risk, raises labor costs
(stocking, facing, ordering, counting), reduces space for true winners, and makes the store harder to shop—without meaningfully improving customer satisfaction or overall profitability.
The operators who consistently outperform treat assortment as a profit-and-capital discipline rather than a volume or variety contest. They ruthlessly protect and expand the 30% that makes the money, keep a tightly controlled middle layer, and systematically eliminate the dead 10–15% (and much of the marginal tail). Everything else is noise that costs more than it returns.
• Store clustering based on real data, not just geography or size. Effective clusters combine POS velocity patterns, census/geodemographic data for the trade area, traffic type, and even competitive set. Two stores in the same zip code can belong to different clusters. • Hybrid assortment models: a true core (the 60–80% of SKUs that belong nearly everywhere) + a meaningful local/flex layer that can differ by store or tight cluster. • Retailer-led data sharing. The chain must push its own POS and demographic insights back to the supplier and insist on store-level or tight-cluster recommendations rather than accepting a regional default. • Technology that makes it practical. Modern assortment optimization and planogram tools (from providers used by both large and mid- sized operators) can generate store- specific or micro-cluster recommendations and planograms at scale. AI-driven systems are particularly good at spotting patterns that humans miss across limited store counts. • Pilot and measure. Test localized mixes in a handful of underperforming stores, track the lift in velocity, margin, and customer feedback, then expand.
Some chains in the 10- 100 store size range have successfully reversed declining stores simply by forcing greater localization—adding or expanding the right ethnic, income-appropriate, daypart, or lifestyle items while aggressively pruning the ones that never belonged there.
The assumption of similarity across a 50–150 mile footprint is frequently false, and it is killing stores. The retailers who push back hardest—armed with their own store-level data—and demand (or build) more granular assortment approaches are the ones that keep those unique locations healthy.
