Designing for efficiency with Systems Thinking

Cut the Complexity: Optimized Inspections in 19 Warehouses

Company

Zalando SE is the leading European online retailer.


In 2024 Zalando:

  • Served 51M+ customers

  • Operated in 25 European markets

  • Delivered 250M+ goods

  • Earned €10.5B in revenue

Team

I was part of a cross-functional team in Zalando's B2B/B2E org.


The team included:

  • 1 Product Manager

  • 1 Program Manager

  • 6 Back end developers

  • 2 Front end developers

Role

As the Leading Product Designer I:

  • Led vision and prioritization workshops

  • Led Jobs-To-Be-Done mapping

  • Conducted user and market research

  • Rapid prototyped wireframes and hi-fi mockups

  • Drove concept and usability tests

  • Supported engineering through implementation

  • Gathered user feedback

  • Continuously iterated

Timeline

Nov 2023 → March 2024

Disclaimer

All product or other sensitive data was anonymized on the screens shared below.

Problem

1. Zalando has a double focus on quality

At Zalando, "quality" means two things:

  1. Meeting EU compliance for product safety: Ensuring all products meet strict European safety regulations.

  2. Setting the platform apart with internal standards: Going beyond compliance to establish rigorous internal benchmarks that differentiate Zalando from competitors.

2. With big success came rapid expansion

In just a few years, Zalando quickly expanded into multiple business models, including Retailer, Marketplace, Discounted Fashion, Fulfillment Solutions, Connected Retail, and more. This rapid growth led to each product having its own set of legal requirements, supply chain processes, and packaging standards.

3. Quality inspections became a bottleneck

To support business growth, Zalando expanded its warehouses, introducing new contracts, tech stacks, system integrations, and workflows. However, these changes created inefficiencies in quality inspection processes, leading to bottlenecks in warehouse operations. Given the importance of quality at Zalando, this raised concerns among leadership.

Goal

I collaborated with a cross-functional team to streamline and design scalable quality inspection tools across Zalando’s warehouse network. Our core challenge was:

How can we ensure consistent, efficient quality checks across all warehouses, without disrupting day-to-day operations?

We identified and defined key performance indicators (KPIs) to target:

100

%

Compliant data stored digitally

Compliant data stored digitally

100

%

Warehouses conducting quality inspections

Warehouses conducting quality inspections

-10

%

Time to complete task

-10

%

Time to complete task

User research

Discoving improvement areas with on-site warehouse visits

I had one month to analyze the workflows of 19 warehouses. Given that I needed to interview at least four teams per warehouse to understand the complex workflows, this timeline was completely unrealistic. Instead, I used a Lean framework.

Running Lean

Focused on two key warehouses

with the most critical operational differences.

Ran collaborative workshops

with cross-functional teams to map out end-to-end flows.

Conducted on-site research

to test assumptions and rapidly iterate solutions.

Validated findings

across a broader set of warehouses to ensure scalability.

Research Findings

Small text caused a big headache

An early key insight revealed that inspections happened far from the computer. Limited visibility caused by small text, cramped interfaces, and cluttered layouts forced users to walk back and forth between inspection sites and their computers. Each inspection required multiple trips!

99% of inspections have no defects

Defects occurred in only 1% of cases but were the most time-consuming part of the inspection process. Each defect report required extra details, photos, a manager review, and and additional handling processes.

It took a long time to find the right item for inspection, and even longer to record the results

Inspectors needed to read identifiers on boxes (such as the shipping number, EAN, or SKU), write them down, and then manually enter them on their computer. Sometimes they’d have to write additional information, like when they identified defects.


Finding the correct items to inspect took 40% of each operator's time when completing an inspection task! Workers had to find specific items of specific SKUs (for example, 20 size L items of Blue shirts in SKU#1). They had to open and sort through multiple boxes to find the items that they needed.

Long German words were causing a headache

Zalando’s warehouses employ people from over 100 different nationalities, meaning German is a second language for many workers. Many struggled to quickly understand the German used in the old inspection software. As a German learner myself, I understood firsthand how challenging it can be to quickly scan and read German, with its long, complex words and intricate grammar.

Design iterations

Testing different ideas for solutions

Falling short with an initial idea

I designed an application to capture notes during ongoing inspections, replicating the existing tools and processes used in two of the warehouses I visited. I quickly ran into concerns about the flow. How would users connect what they wrote in the application to the physical goods? And wouldn't an approach like this just recreate the too-many extra manual steps that existed in the current processes?

Thinking hard about hardware

Thinking about the prevalance of barcodes in the warehouses, I explored another version where the user began the flow by scanning the item's label, which quickly evolved into using barcodes for all interactions. I also ran into several questions:

  • What’s the most suitable hardware for interaction—scanner, tablet, phone, or POS keyboard?

  • And how can we minimize the number of interactions so that user don't need to scan too many labels?


I explored and explored. I ended up trying the different versions onsite to test which ones would be the most effective. I brought paper prototype and we role played the different concept directly at the work-stations with heads of departments and end users.

Solution

Cutting inspection time from minutes to seconds: 99% done in two scans

1. We integrated digital and physical tools

The solution ultimately consisted of :

  1. a new application

  2. a barcode scanner

  3. a physical barcode book.


The app made it easy to input data and work smoothly with the barcode scanner. The physical barcode book was a familiar tool for workers, making the transition easy and allowing them to start using it right away.

2. Barcodes were the key to simplifying the process

We embraced the barcode scanner because it didn’t require additional security or login flows, unlike phones or tablets. It was portable, familiar to users, and already deployed across all warehouses, which meant it didn't need leadership or union approval. Each worker also received a barcode book, similar to those used at supermarket checkouts.


Using barcodes meant that inspecting an item was a simple 2 step process:

  1. Users scanned the barcodes on the boxes, palettes or packages

  2. Users communicated the result of the inspection by scanning barcodes in their booklets

3. 99% of inspections were finished in 2 scans

This streamlined the process for non-defective inspections, which accounted for 99% of cases. After just two scans, inspection work was completed, eliminating the need for inspectors to manually enter shipment numbers, update status information, or return to their workstations.


When an item required inspection, the system instantly alerted the inspector upon scanning, so there was no longer a need to search through boxes in search of the correct items.

4. The remaining 1% of inspections
were completed in 4 steps

Reporting the defective items took the most time, so I aimed to simplify the process by designing a straightforward 4-step flow:

  1. Scan the barcode on the package

  2. Scan the error code

  3. Scan the error level

  4. Take a picture of the issue

5. A cleaner interface made it easier to use from afar

Finally, I simplified the interface by displaying only the most critical features and introducing smart defaults. I also increased the size of key information to enhance readability. Additionally, workers could opt to use the tool in English, one of the most commonly spoken languages in the warehouse.

This allowed users to navigate the workflow even from a distance, without needing to stand directly in front of their computer.

Impact

Design with Results: Reducing Time and Costs Across Operations

100

%

Compliant data stored digitally


100% of compliance data was stored and available in case of audits. Avoiding a 2% fine on Zalando yearly revenue.

100

%

Compliant data stored digitally


100% of compliance data was stored and available in case of audits. Avoiding a 2% fine on Zalando yearly revenue.

100

%

Compliant data stored digitally


100% of compliance data was stored and available in case of audits. Avoiding a 2% fine on Zalando yearly revenue.

100

%

Warehouses conducting quality inspections

By December 2024, three warehouses had been successfully onboarded. Following positive feedback from department heads and noticeable efficiency gains, a full rollout to all 19 warehouses is scheduled for completion by the end of 2025.

-20

%

Time to complete task


I reduced the time to complete a quality inspection by 20%. Which is a huge win!

-20

%

Time to complete task


I reduced the time to complete a quality inspection by 20%. Which is a huge win!

-20

%

Time to complete task


I reduced the time to complete a quality inspection by 20%. Which is a huge win!

© Tristan Bézard 2025