Learn what On-Shelf Availability (OSA) means in retail, how to calculate it, ideal benchmarks (95–98%), common causes of stockouts, and proven strategies to improve shelf availability and protect revenue.
This guide explains how to calculate OSA, choose a meaningful benchmark, distinguish shelf out-of-stocks from true store stockouts, diagnose the root cause of shelf gaps, and build a closed-loop process to improve availability.
On-shelf availability measures whether a product is physically available and accessible to the shopper at its expected selling location when the shopper wants to buy it.
The key word is shopper.
Inventory availability asks:
“Does the retailer have the product?”
OSA asks:
“Can the shopper find and buy the product?”
Those questions are not always answered the same way.
Consider this example:
The retailer technically has stock. But from the shopper’s perspective, the product is unavailable.
This distinction is central to OSA. ECR Retail Loss defines the problem from the shopper’s perspective: whether the shopper can find the product on the shelf when they want it. It also distinguishes this from simply having stock somewhere inside the store.
OSA therefore sits at the end of a much longer retail chain:
Demand → Supply → Store Inventory → Backroom → Replenishment → Shelf Execution → Shopper Access → Sale
A failure anywhere along this chain can eventually appear as an empty shelf.
That is why an empty shelf should not automatically be treated as a supply problem.
OSA measures shopper-accessible product availability, not simply whether inventory exists somewhere in the retail network.
OSA is calculated by dividing the number of expected SKUs available on the shelf by the total number of expected SKUs, then multiplying by 100.
The formula is easy. Defining the denominator correctly is harder.
The total should represent the products that were actually expected to be available during the measurement period.
Depending on the purpose of the audit, a retailer might use:
For example, suppose a retailer has 1,000 products in its master catalogue, but Store A is authorized to carry only 600. If 570 of those 600 are available, its operational OSA is 95%, not 57%.
A useful OSA report should therefore document at least three things:
Without those rules, two teams can inspect the same stores and produce different OSA numbers.
Key takeaway: The OSA formula is simple, but the result is meaningful only when the retailer has a clearly defined SKU universe, observation period, and availability rule.
OSA measures shopper availability; store OOS and shelf OOS describe different types of product absence; inventory availability describes what the retailer believes or knows is in stock.
This distinction matters because the corrective action changes with the diagnosis.
ECR research specifically separates store out-of-stocks from shelf out-of-stocks. A shelf OOS can occur when the product is in the backroom, in the wrong location, in a cage or trolley, or otherwise unavailable to the shopper despite being somewhere on the premises.
Consider:
System inventory: 12
Backroom inventory: 12
Shelf inventory: 0
Ordering another 12 units is unlikely to solve the immediate problem.
The first action is to move existing stock from the backroom to the selling shelf.
Now consider:
System inventory: 0
Backroom inventory: 0
Shelf inventory: 0
That requires a different investigation: supply, ordering, allocation, distribution or replenishment parameters.
OSA tells you what the shopper experiences. OOS and inventory information help explain where the availability chain failed.
The fastest way to diagnose OSA is to trace the shelf gap through five layers: Supply → Inventory → Backroom → Shelf Execution → Shopper Availability.
An empty shelf is an observation, not a root cause.
Start upstream.
Ask:
Was the expected product actually delivered to the store?
Possible causes include:
If the product never arrived, a store associate cannot replenish it.
Next ask:
Can the retailer trust the inventory record?
Inventory records can become inaccurate through receiving errors, incorrect adjustments, theft, damage, misplaced products, checkout errors and other process failures.
This is not a theoretical issue. ECR research found that correcting inventory inaccuracies produced approximately 4–8% sales increases across participating retailers in controlled test-and-control experiments. The research also found that inaccurate records can lead to stock being replenished too late or too early.
A system showing 10 units does not prove that 10 saleable units are actually available.
If physical inventory exists, ask:
Is saleable stock sitting in the backroom while the shelf is empty?
This is one of the clearest examples of a shelf OOS.
The store has inventory. The shopper does not.
The operational causes may include poor backroom organization, inconsistent shelf checks, unclear ownership, insufficient labor or weak replenishment routines.
If the product has reached the sales floor, ask:
Is it in the correct location, with the required facings and visibility?
Check:
A product can physically exist on the sales floor and still fail the intended shelf standard.
The final test is simple:
Can the shopper find, reach and purchase the product when they want it?
That is the outcome OSA is ultimately trying to measure.
There is no single OSA percentage that should be treated as the right target for every retailer, category and SKU.
The old Pazo article presented 95–98% as a general strong OSA range and gave category-specific ranges. Those figures are too broad to use as universal targets without defining the measurement method and business context.
A better approach is to set targets according to:
Category × SKU velocity × Store format × Replenishment model × Promotion × Economics
For example, a fast-moving beverage SKU during a major promotion may deserve a more aggressive availability target than a slow-moving product with limited shelf life.
The economics also matter.
In fresh and perishable categories, pushing availability higher can require more inventory and can increase waste. ECR research explicitly describes OSA and waste as a trade-off and notes that the appropriate target can differ between dry grocery and perishables.
Published benchmarks can provide context, but they should not automatically become operational targets. APQC currently reports a median 88.0% for its cross-industry retail on-shelf/in-stock availability measure across 350 companies; importantly, its measure covers both online and store operations, so it should not be interpreted as a universal physical-store OSA target.
Track OSA by:
This tells you whether the problem is broad or concentrated.
A useful OSA target is not simply the highest possible percentage. It is the availability level that protects important sales while balancing inventory, labor, replenishment and waste economics.
Retailers should measure OSA at SKU-store level and then aggregate the results to identify patterns by store, category, region, SKU and time period.
A store-wide OSA percentage is useful, but it can hide important failures.
Establish which SKUs the store should have available.
Do not automatically include discontinued, inactive or unauthorized products.
Set a consistent rule.
For example:
A product is available when saleable stock is physically accessible to shoppers at its intended selling location.
The rule should be applied consistently across auditors and stores.
Measurement can come from:
ECR identifies independent audits, internal store audits and distribution/service-level measures among approaches used to understand shelf out-of-stocks.
Track:
Store → Category → SKU → Priority SKU → Promotion → Time
The last three are particularly important.
A store may report 95% OSA while the five missing products are its five highest-volume SKUs.
That is a very different problem from losing five slow-moving products.
A shelf gap observed once is not equivalent to a SKU that repeatedly disappears for several hours.
Track:
This changes OSA from a static score into an operational performance measure.
Good OSA measurement answers three questions: How much availability are we achieving? Where are we losing it? Why do the same failures keep recurring?
Retailers improve OSA by connecting shelf detection to root-cause diagnosis, corrective action, verification and prevention.
A practical operating loop is:
Detect → Diagnose → Assign → Correct → Verify → Prevent
Identify which products are unavailable and where the failures are concentrated.
Prioritize:
Not every SKU carries the same commercial risk.
Ask:
Was it supplied?
Is inventory accurate?
Is stock in the backroom?
Was it replenished?
Is it correctly positioned?
Is shelf capacity sufficient?
This prevents a retailer from treating every shelf gap as a purchasing problem.
An OSA observation should create a clear operational responsibility.
A useful corrective task contains:
For example:
SKU: Product A
Issue: Shelf OOS
Root cause: 8 units in backroom
Action: Replenish shelf
Owner: Store team
Verification: Shelf photograph
The observation has now become an executable task.
A task marked “complete” does not necessarily mean the shelf is fixed.
Verification can include:
If the same SKU repeatedly becomes unavailable, investigate the pattern.
The underlying issue may be:
The objective is not simply to close more tasks. It is to reduce the number of repeat shelf gaps.
Planogram compliance and OSA measure different things, but poor shelf execution can directly affect product availability, visibility and replenishment.
OSA asks:
Can the shopper find and buy the product?
Planogram compliance asks:
Is the shelf being executed according to the intended merchandising standard?
Consider a product that should have three facings.
If the store gives it only one facing because another SKU has taken the space, the product may still be available—but its shelf capacity and visibility have changed.
Now consider the reverse:
The planogram may be compliant while OSA is poor.
That is why retailers should analyze OSA alongside:
The relationship is especially important for high-velocity SKUs. A product with limited facings may sell through between replenishment cycles even though the store has adequate total inventory.
Planogram problem: The shelf is not built or maintained according to the intended standard.
OSA problem: The shopper cannot obtain the expected product.
Combined problem: Poor shelf design or execution is contributing to repeated availability failures.
Ultimately, improving OSA is not about keeping more inventory everywhere. It is about making sure the right product reaches the right shelf, at the right time, and remains available to the shopper when the purchase decision happens.
As someone who has built highly scalable products from the ground up, I've always been drawn to solving challenging problems. But it's the quest for operational excellence that truly lights my fire. The thrill of streamlining processes, optimizing efficiency, and bringing out the best in a business – that's what gets me out of bed in the morning. Whether I'm knee-deep in programming or strategizing solutions, my focus is on creating a ripple effect of excellence that transforms not just businesses, but the industry at large. Ready to join forces and raise the bar for operational excellence? Let's connect and make retail operations and Facilities Management better, together.
Stay up to date with the latest video business news, strategies, and insights sent straight to your inbox!
