Shelf fill rate is the proportion of available shelf facings occupied by product at a given measurement point, or averaged across a defined time window. The common flaw is treating one point-in-time count, usually taken at shift start, as if it were the store's fill rate for the whole day.
That distinction matters. A shelf can look well stocked at 08:00 and show repeated gaps during the busiest trading hours. If the morning snapshot is the only measurement, the calculation will overstate availability and may lead to restocking schedules that are too late or too infrequent.
Start with the facing count
A facing is a single product unit visible from the customer aisle, based on the width of the product as displayed. To calculate fill rate, first define the shelf area and count the facings that should be available there.
The basic point-in-time formula is:
Fill rate = occupied facings ÷ available facings × 100
For example, if a shelf section has 40 available facings and 34 are occupied when checked, its fill rate is 85%.
The calculation needs a consistent definition of “available.” If a bay is intended to hold 40 facings, use 40 as the denominator even when the shelf is temporarily empty. If a facing has been deliberately removed as part of a range or merchandising change, update the expected facing count before comparing results.
It is also useful to record the measurement area. A whole-store figure can conceal problems in a particular gondola bay, chilled section, or high-volume category.
Point-in-time versus time-averaged fill rate
A point-in-time rate answers a narrow question: how full was the visible shelf when someone looked? A time-averaged rate answers a different operational question: how full was the shelf, on average, during the trading period?
To calculate a time-averaged rate, take repeated measurements at defined intervals and average them, or calculate the occupied facings across the full set of observations.
Consider a zone with 40 available facings:
| Time | Occupied facings | Fill rate |
|---|---|---|
| 08:00 | 38 | 95% |
| 12:00 | 35 | 87.5% |
| 15:00 | 27 | 67.5% |
| 18:00 | 32 | 80% |
The shift-start result is 95%. The simple average of these observations is 82.5%. If the 15:00 gap is a recurring pattern, the second figure is more useful for deciding when to schedule a replenishment check.
Where observation intervals are not equal, use a weighted calculation. A two-hour period with a 70% fill rate should not count the same as a 15-minute period with a 95% fill rate if the goal is to describe the full trading window. In practice, define the measurement interval first, then apply the same interval across stores and zones.
Zone-level and category-level measurement
Zone-level fill rate is calculated for a defined section of shelf, such as one gondola bay or a product category zone. It is usually the most useful level for daily action because it tells a team where to send someone.
Category-level fill rate covers all SKUs in a product category. For example, a drinks category may include several bays, package sizes, and brands. A category figure can show whether the category is generally available, but it can also hide a weak zone if other shelves are full.
Use the level that matches the decision:
- Use zone-level fill rate to set restock priorities and identify recurring problem areas.
- Use category-level fill rate to compare broad availability between stores or trading periods.
- Use store-level results for area reviews, while retaining the underlying zone data for follow-up.
When comparing categories, avoid giving a small category and a large category equal weight without checking the facing counts. A category with 20 facings and a category with 200 facings do not contribute the same amount of shelf capacity. For a broader store calculation, add occupied facings across the relevant zones and divide by the total available facings.
How manual measurement works, and where it breaks down
The manual method is straightforward. An employee walks the store at a scheduled time, counts occupied and available facings in selected zones, records the result, and repeats the process later. This can work for a small number of stores or for a targeted audit.
The limitation is coverage. A single shift-start walk does not show what happens after a delivery, a promotion, or a busy customer period. Even several checks can miss a short gap window. Results can also vary between people if the zone boundaries, expected facings, or treatment of partially empty shelves have not been defined clearly.
One anonymous example is a 12-store convenience chain. Its shift-start reports showed fill rates above the internal target in most stores. After the chain compared observations throughout the day, it found a consistent 14:00 to 16:00 decline in several high-traffic zones. The shelves were replenished before opening, but the schedule did not account for the afternoon gap window. The operational response was not to add a general stock count. It was to move a replenishment check into that period and focus it on the affected zones.
Camera-based monitoring can provide more frequent visibility from existing ONVIF cameras. Youtiligent computes zone-level fill rate per shift from camera feeds and alerts when a configured fill threshold is crossed, rather than waiting for the next manual check. Alert sensitivity can be set to high, medium, or low. Shelf fill monitoring is included in the Pilot tier for 1 to 5 stores at $349 per month and in all higher tiers. Processing is performed on-premises, so video footage does not leave the network.
Important measurement boundaries
Camera fill rate inference is not perfect for every shelf configuration. Very low shelves may not be fully visible from the camera angle. Shelving behind pillars can be obstructed, and narrow aisles may limit where a camera can be placed. Reflections, product presentation, and changes to the shelf layout can also affect what is visible.
The result measures visible shelf occupancy, not stock in the back room. It does not replace physical stock counts for inventory management, receiving, ordering, or reconciliation. Use it as an availability and restock signal, then investigate the cause of a low reading through the normal store process.
For a practical operating standard, define the available facings, divide the store into named zones, choose a measurement interval, and report both shift-start and time-averaged results. If the time-averaged rate falls below the target during a repeated trading window, schedule the restock check inside that window instead of relying on the opening count.
Write to the Youtiligent team if you want to discuss zone-level shelf fill monitoring for your stores.
Back to Store Ops Insights