Picking performance: the KPIs that signal it’s time to change your system

When a picking system stops being sufficient, it rarely does so abruptly. What usually happens is that performance indicators gradually deteriorate, each one seeming manageable on its own, and the real problem only becomes visible when they are analyzed together.

The challenge for any operations director or logistics manager is to identify that point before the system becomes a bottleneck to growth. And the only way to do this rigorously is by measuring. Not managing by intuition, but by data.

For over 25 years, we have helped pharmaceutical distribution, retail, and e-commerce companies analyze their picking operations and implement systems that solve problems at their root. What we outline below are the indicators that, in our experience, most clearly signal when change is no longer optional.

Signs that tend to appear together

Before looking at each KPI individually, it’s important to recognize the pattern. In practice, it’s rare for only one of these indicators to appear. More commonly, several deteriorate at the same time, confirming that the issue is systemic rather than isolated:

• Lines per hour stagnating or declining without changes in order volume or complexity
• Error rates remaining high without a clear operational cause
• Cost per order steadily increasing without business justification
• Demand peaks repeatedly disrupting operations
• Onboarding time for new staff becoming longer than expected

Each of these on its own may seem manageable. Together, they point to a structural problem. And structural problems cannot be solved with partial adjustments.

KPI 1: Lines per hour per operator

This is the most direct productivity indicator. It measures how many order lines each operator completes per hour of effective work, excluding breaks and transition times.

Typical manual operations range between 80 and 120 lines per hour, depending on order type and warehouse layout. Facilities using assisted picking technologies can double that performance under comparable conditions, with the same workforce. If your operation has been stuck in the lower range for months without justification from volume or complexity, the system has a structural issue that no staffing adjustment will fix.

The most common cause is travel time. In unguided manual operations, operators can spend between 60% and 70% of their shift moving around the warehouse—time that generates no order lines. Introducing guided systems that allow multiple orders to be picked in a single optimized route is typically the change with the greatest immediate impact on this KPI.

KPI 2: Error rate per line

Manual picking without technological assistance typically has an error rate between 1% and 3% per line. With guided systems, this figure consistently drops below 0.5%, depending on SKU type and verification processes.

The difference may seem small until translated into real volume. In an operation handling 2,000 lines per day, moving from 2% to 0.5% means going from 40 daily errors to 10. Each error carries a direct cost: returns handling, order replacement, customer service time, and in regulated sectors such as pharmaceuticals, potential impact on traceability and compliance.

When errors persist without a clear operational cause, it is usually because the system forces operators to make decisions at every step instead of guiding them. Each unguided decision is an opportunity for error. Visual guidance systems at the pick point eliminate this margin: the operator receives exact instructions on what to pick and in what quantity, with built-in confirmation before moving on.

KPI 3: Cost per order

Cost per order is the KPI that connects operational efficiency with financial performance. It includes labor costs, the cost of errors and returns, and the cost of system inefficiencies as a whole.

In many cases, this KPI is not measured directly, but it shows up through clear symptoms: more hours required for the same volume, recurring incidents, or the need to reinforce the team without proportional business growth.

When it increases steadily without clear explanation, the system is consuming more resources to produce the same output. Operations that seriously optimize their picking processes report reductions in operational time of between 25% and 45% within the first months—a wide range depending on the starting point and the solution implemented. If your cost per order has been rising quarter after quarter without a clear reason, it’s not a business issue. It’s a system issue.

KPI 4: Scalability during demand peaks

This is the most underestimated KPI under normal conditions—and the most painful when a peak arrives. The key question is not whether the system can handle regular volume, but how much additional volume it can absorb without losing operational control.

A system that needs to double the workforce to handle a 30% increase in volume is not scalable. A well-designed system should absorb that increase with the same resources or a proportionally smaller adjustment. If every major campaign requires structural improvisation, the system has a scalability limit that is constraining business growth. In that scenario, the team works faster—but with less control—and that difference shows up later in errors and returns.

KPI 5: Onboarding time for new staff

This is a less obvious but highly revealing KPI. It measures how long it takes for a new operator to reach the team’s standard performance level.

When this time is long, it usually indicates that the system depends on tacit knowledge held by experienced staff rather than on a clear, guided process. This creates dependency on specific individuals, vulnerability to turnover, and difficulty scaling the team during peak periods without performance loss.

A visual guidance system solves this at its core. The operator does not need to memorize locations or procedures: instructions at the pick point clearly indicate what to do at every step. In operations with high turnover or frequent temporary staff, this factor can be as decisive as error rate or cost per order.

How to justify the change internally

When several of these KPIs show sustained deterioration at the same time, the next step is to build a return-on-investment analysis to justify the change. The key elements are the current cost of errors and returns, the cost of unproductive travel time, the cost of dependency on specialized staff, and the economic impact of poorly managed demand peaks.

Investment in a technology-guided picking system delivers a concrete and measurable return. Integration with the existing WMS is typically done via CSV or XML files in batch mode, TCP/IP protocol for real-time communication, or standard libraries for companies with in-house IT capabilities. This ensures that the new system can be incorporated into the existing infrastructure without replacing it. In most cases, the transition is carried out in phases, starting with high-rotation areas, allowing performance to be validated under real conditions before scaling.

If you recognize these symptoms in your operation, the issue is likely not your team—but your system. Analyzing KPIs with real data is the first step to understanding whether change makes sense and how quickly it will pay off. If needed, we can help you assess your operation before making any decision. You can explore our guided picking solutions or contact us directly.

FAQs

Which KPIs should I measure to know if my picking system has reached its limit?
The most relevant are lines per hour per operator, error rate per line, cost per order, scalability during demand peaks, and onboarding time for new staff. Analyzed together, they provide a clear picture of system performance and real improvement potential. If you’re not measuring one of them, that blind spot is often where the most costly issue lies.

What is an acceptable picking error rate?
Manual picking without technological assistance typically ranges between 1% and 3% error per line. With guided systems, this drops below 0.5%. In regulated sectors like pharmaceuticals, where traceability is mandatory, this margin is not negotiable.

How does a guided picking system integrate with an existing WMS?
Integration is usually done via CSV or XML files in batch mode, TCP/IP protocol for real-time communication, or libraries in .NET, C++, and Java for companies with internal IT resources. For companies without technical resources, specialized integrators can manage the full implementation process.

How long does it take to achieve ROI?
It depends on volume, current error rate, and labor costs. In medium to high-volume operations, payback typically occurs within 18 to 36 months, mainly driven by reduced errors and returns, and increased lines per hour without increasing staff. A prior analysis allows for a more precise estimate based on real data.

Can the system be implemented in phases without disrupting operations?
Yes. The usual approach is to start with high-rotation areas, validate performance under real conditions, and progressively extend the solution. This avoids disruption and allows adjustments before scaling.

What’s the difference between fixed pick-to-light systems and carts with integrated displays?
Fixed pick-to-light systems are installed on shelving and indicate what to pick at each location. Carts with integrated displays are mobile solutions that carry guidance with them, allowing multiple orders to be picked in a single route without fixed infrastructure. Both can be complementary depending on the operation, and the choice depends on volume, number of SKUs, and warehouse layout.