SLA (Service Level Agreement): What to Include in Your 3PL Contract
09.12.2025
Inventory Shrinkage: Why Stock Disappears and How to Prevent It
09.12.2025
OUR GOAL
To provide an A-to-Z e-commerce logistics solution that would complete Amazon fulfillment network in the European Union.
Imagine looking at your monthly logistics dashboard. The report glows green: "Inventory Performance: 95%." It looks like a win. Yet, your Customer Support ticket queue is overflowing with complaints about delayed shipments, and your marketing team is furious about "Out of Stock" labels on your bestsellers during a campaign.
How can the metrics look so good while the customer experience feels so disjointed?
This is the classic trap of confusing Fill Rate with Service Level. While often used interchangeably in casual boardroom conversations, these two metrics measure completely different aspects of your supply chain health. One looks backward at what you achieved; the other looks forward at what you are prepared to handle.
For an e-commerce manager or operations director, understanding the nuance between these two is not just about semantic pedantryāit is the difference between carrying excessive dead stock and losing your most loyal customers to a competitor with better availability.
Fill rate vs. Service level: Why we mix them up
At a high level, both metrics attempt to answer the same question: "Did we have the product when the customer wanted it?"
However, the perspective differs radically.
- Fill rate is an execution metric. It measures immediate gratificationāthe percentage of customer demand met without delay from available stock.
- Service level is a planning metric (specifically, Cycle Service Level). It is a probability calculation used to determine safety stock. It measures the likelihood that you will not face a stockout during a specific replenishment cycle.
In simpler terms: Fill Rate tells you how many orders you successfully shipped yesterday. Service Level tells you how confident you are that you won't run out of stock before the next truck arrives.
Decoding fill rate in e-commerce
Fill rate is the most tangible metric for e-commerce because it directly correlates to revenue capture. If a customer tries to buy an item and you ship it immediately, your fill rate is 100% for that transaction. But in the complex world of multi-line orders, "Fill Rate" splits into three distinct KPIs.
1. Unit fill rate (Volume metric)
This measures the percentage of individual items ordered that were shipped immediately.
- Formula: (Total Units Shipped / Total Units Ordered) x 100
2. Line fill rate (SKU metric)
This measures the percentage of order lines (specific SKUs) that were completely fulfilled. If a customer orders 10 units of SKU A, and you ship 9, the Line Fill for that line is 0% (depending on your strictness) or 90% (volume based). Usually, in logistics, a line is considered "filled" only if the specific request is fully met.
3. Order fill rate (Customer experience metric)
This is the most brutal, yet most honest metric for e-commerce. It measures the percentage of complete orders shipped without splitting shipments or backorders.
- Scenario: A customer orders a Shampoo, a Conditioner, and a Hair Mask. You have the Shampoo and Conditioner, but the Mask is out of stock.
- Unit fill: 66%
- Order fill: 0%
Why this matters: If you rely on Unit Fill or Line Fill, you might think you are performing at 98%. But if your Order Fill is only 85%, that means 15% of your customers are experiencing a delay, a split shipment (which doubles your shipping costs), or a cancellation.
Statistical reality of service level (CSL)
While Fill Rate is empirical (you can count it), Service Level is probabilistic. It is deeply rooted in inventory management and safety stock calculations.
When a supply chain manager says, "We target a 95% Service Level," they are not saying they will fulfill 95% of orders. They are saying: "There is a 95% statistical probability that we will not face a stockout during the lead time of replenishment."
This metric deals with variabilityāboth in demand (spikes in sales) and supply (delays from manufacturers).
Math behind the curtain
To calculate the required inventory for a specific Service Level, you must use the standard normal distribution (the bell curve). The formula for Safety Stock involves:
- Z-score: The coefficient corresponding to your target service level (e.g., 1.65 for 95%, 2.33 for 99%).
- Standard deviation of demand: How much your sales fluctuate.
- Lead time: How long it takes to restock.
"Diminishing returns" trap
This is where Service Level differs most from Fill Rate. Increasing your Service Level from 95% to 99% does not require a 4% increase in inventory. Because of the non-linear nature of the bell curve, it might require doubling your safety stock.
For an e-commerce business, aiming for a blanket 99% Service Level across the entire catalog is usually financial suicide. It ties up massive amounts of cash in working capital for slow-moving items.
Comparative analysis: Differences at a glance
To visualize the distinction, letās look at a specific scenario.
Scenario: You sell premium coffee machines.
- Demand: You expect to sell 100 units this week.
- Actual demand: 110 customers try to buy them.
Stock: You have 100 units on hand.
Outcome:
You sell all 100 units. You lose 10 sales.
- Fill rate: 90.9% (100 units shipped / 110 demanded). You satisfied roughly 91% of the demand.
- Service level: 0%. Why? because a stockout occurred. You did not survive the cycle without running out. In the binary view of Cycle Service Level (Stockout vs. No Stockout), you failed the cycle.
| Ā | Fill Rate | Service Level (Cycle) |
Perspective | Retrospective (What happened?) | Prospective (What are we planning for?) |
Focus | Sales & Order Fulfillment | Inventory Planning & Safety Stock |
Calculation | Simple Arithmetic (Shipped / Ordered) | Statistical Probability (Z-scores) |
Impact | Direct Revenue & Customer Satisfaction | Capital Investment & Holding Costs |
Best Used For | Measuring Operational Performance | Setting Inventory Targets |
Financial trade-off: Inventory holding cost vs. cost of lost sales
In e-commerce logistics, the tension between Fill Rate and Service Level is essentially a tension between two costs:
- The cost of having too much stock (high service level): Warehousing fees, obsolescence risk, tied-up cash flow.
- The cost of having too little stock (low fill rate): Lost margin, damaged brand reputation, increased Customer Acquisition Cost (CAC) because you have to "re-acquire" a disappointed customer.
"Amazon Effect" on expectations
Consumer expectations have shifted the goalposts. Years ago, a backorder notification was acceptable. Today, if an item is not available for immediate dispatch, the cart is abandoned. This puts immense pressure on Fill Rates.
However, if you blindly chase a 100% Fill Rate, you will overstock. The solution lies in Inventory Segmentation.
Applying the ABC analysis
You should not treat a fast-moving, high-margin SKU the same way you treat a slow-moving accessory. Expert inventory managers decouple their Service Level targets based on the SKU profile.
Class A items (High value / high velocity)
- Strategy: Aggressive.
- Target: 98-99% Service Level.
- Why: These are your bread and butter. A stockout here hurts revenue immediately. You justify the higher safety stock cost because the turnover is fast.
Class B items (Moderate value / moderate velocity)
- Strategy: Balanced.
- Target: 90-95% Service Level.
- Why: Customers might be willing to wait a day or two, or substitute the product.
Class C items (Low value / low velocity)
- Strategy: Conservative.
- Target: 85-90% Service Level.
- Why: Holding 6 months of inventory for an item that sells twice a month is inefficient. For these items, a lower fill rate is an acceptable trade-off for cash flow health.
OTIF: Modern metric bridging the gap
In modern logistics (especially when working with 3PL partners), a third metric often enters the conversation to bridge the gap between planning and execution: OTIF (On-Time In-Full).
While Fill Rate measures if you shipped it, OTIF measures if you shipped the correct amount at the correct time.
- On-time: Did the warehouse process the order within the SLA (e.g., same-day dispatch before 14:00)?
- In-full: Was the order perfect, or were lines missing?
For e-commerce, OTIF is the closest proxy to "Customer Happiness." A 99% Fill Rate means nothing if the warehouse takes 4 days to pick the pack. Tracking OTIF ensures that your high Service Level planning actually translates into a superior customer experience.
How to improve fill rates without inflating inventory
If your Fill Rate is suffering but you cannot afford to increase your Service Level (and thus your stock levels), the issue is likely not the inventory quantity, but the inventory quality or data accuracy.
1. Fix phantom inventory
One of the biggest killers of Fill Rate is "Phantom Inventory"āwhen your system says you have 5 units, but the bin is empty. This leads to accepted orders that cannot be fulfilled. Regular cycle counting (rather than just annual audits) is the cure.
2. Vendor lead time reduction
The safety stock formula relies heavily on Lead Time. If you can negotiate faster restocking with your suppliers (or switch to local suppliers), you can maintain the same Service Level with significantly less stock.
3. Demand forecasting integration
Are you running marketing campaigns without informing your logistics team? If Marketing pushes a "20% Off" email, demand spikes. If the Service Level was calculated on historical averages, you will stock out. Integrated planning (S&OP) aligns the Service Level targets with upcoming promotional calendars.
Moving beyond the percentages
Ultimately, Fill Rate and Service Level are not just numbers on a spreadsheet; they are the pulse of your supply chain strategy. Service Level is the promise you make to your business about how much risk you are willing to tolerate. Fill Rate is the reality check of how well you kept that promise to your customer.
For growing e-commerce brands, the goal isn't to maximize both metrics indiscriminately. It is to find the "Sweet Spot"āthe point where the cost of inventory intersects optimally with the cost of lost sales.
Navigating this trade-off requires more than just good software; it requires agile logistics execution. Whether you manage your own warehouse or partner with a 3PL, the ability to react to data discrepancies and adjust safety stocks dynamically is what separates fragile online shops from resilient e-commerce leaders. The next time you look at that dashboard, look past the green "95%" and ask: is this the right 95% for my bottom line?
