If you’ve ever shopped for fashion online, you know the feeling: endless scrolling, a cart full of potential fits, and that crucial moment right after you hit “Buy now”.

At Zalando, we serve over 62 million active customers across 29 European markets. With a design community of over 130 practitioners, our job isn’t just making things look good — it’s creating seamless, joyful experiences across the entire purchase journey.

Recently, Emrah Sarıbaz, Head of Product Design, and Vedanti Joshi, Principal Product Designer, took the stage at WaysConf to share how our team translated insights into measurable impact. They unpacked how we built a fast-moving, high-impact experimentation space to test our way to scale.

Here is how we shifted our mindset, delivered our goals ahead of schedule, and learned a few surprising lessons along the way.

The challenge: An ambitious goal

It all started with a clear challenge from senior leadership: generate a gross merchandise value (GMV) uplift within a specific timeframe.

In e-commerce, GMV is an output metric. It’s the top of the mountain. But as product designers working on everything after the “decide” moment — bag, checkout, payments, delivery tracking, and returns — we influence the input metrics that drive that output:

• Conversion rates
• Return rates
• Customer care contacts
• Average order value
• Customer engagement

Our traditional product development process follows four stages: Problem Definition → Solution Design → Execution → Post-Launch. While fantastic for complex feature rollouts, dependencies, app release cycles and alignments mean shipping value through this pipeline takes a few weeks per initiative. Moreover, our cross-functional product trios (Design, PM, Engineering, Analytics) were already fully allocated to portfolio initiatives in our roadmap.

To hit our target rapidly without disrupting core roadmaps, we requested a dedicated budget to form a nimble task force and test a brand-new way of working.

The “secret sauce”

Spoiler alert: budget wasn’t the secret sauce. It just unlocked the room. The real secret sauce was stripping away friction and changing our team mindset.

1. Shared ownership & guiding principles

We empowered the task force with full ownership. Clear ownership was the cornerstone of the success with streamlined decision making. Anyone on the team could unblock the team’s progress as we defined the guiding principles to further simplify the decision making and steer the ship on the right direction. Our core principles like “deliver with low maintenance in mind” and “play to win, dare to fail” helped us to focus on what matters the most.

2. Tool simplicity

We skipped complex software setups and prioritised making the process seamless, visible and scalable. We implemented a experimentation design document template, a Google doc, to store all the necessary information such as hypotheses, KPIs, platform or markets in a single and repeatable format, then created a roadmap on Google sheet to consolidate all completed, ongoing, prioritised and backlog hypotheses with clear status updates and timeline, and finally built a single Figma file as the source of truth for implementation handoff and content design.

3. Embracing the power law

We bypassed lengthy ideation workshops and went straight to a goldmine: 50+ high-potential backlog insights from prior research that had never been prioritised. Under the Power Law (similar to the 80/20 Pareto principle), experimentation outcomes are heavily skewed, most lift is generated by a small fraction of bold tests. Knowing we couldn’t predict every winner upfront, we focused on test velocity. By increasing our throughput, we gave ourselves more shots on goal to surface those rare, high-leverage wins.

Ranking ideas with RICE

We mapped frequently used prioritization frameworks across two axes — Speed vs. Rigour and Data-Backed vs. Gut Feeling — giving us a clear picture of how each works so we could sequence our experiments effectively.

While ICE and PIE are great quick-checks, we leaned heavily into RICE (Reach, Impact, Confidence, Effort). RICE forces a concrete numeric definition on effort — estimating exact person-hours and dev time — converting fuzzy timeline assumptions into hard data.

Once scored, every idea entered a continuous loop: Score Test Evaluate Scale, Retire, or Rescore.

From ideas to experiments: What worked, what didn’t and what stung

Like the famous Sony Walkman focus group — where everyone requested yellow in interviews but took the black one home — user words don’t always match live behaviour. Here is what happened when we put some cart and checkout hypotheses in front of millions of users:

Adding a “Highly rated” social proof tag

• The idea: Adding a subtle “Highly rated” badge to items in the cart rated 4.5+ stars.
• Result: Slightly positive! Timing is key for reassurance messaging: it drives conversions at the point of decision, but feels like clutter if introduced too late or too early.

Streamlining the empty bag page

• The idea: Removing static illustrations from empty cart pages and moving the recommendation carousels higher.
• Result: Positive engagement lift! An empty cart isn’t a lost customer; direct guidance beats decorative visual distraction.

“Low in Stock” scarcity messaging

• The idea: Displaying real-time stock scarcity badges on bag items.
• Result: Massive GMV uplift! This single test drove the highest lift in our portfolio. It didn’t invent demand — it collapsed the decision window by 70%, encouraging buyers to complete checkout faster.

Free delivery progress bar (A/B test)

• The idea: Showing dynamic banners explaining how close customers were to free shipping thresholds.
• Result: Near-flat conversion. While usability click tests looked promising, live A/B tests proved users tuned out the banner completely. “The more we said, the less they heard.”

The “Woohoo!” celebration moment

• The idea: A delightful post-checkout celebration state.
• Result: No-harm experiment! Financial metrics remained flat, but qualitative brand perception soared in key markets — giving us valuable insights into regional sentiment and localization needs.

The final scoreboard

By running rapid experimentation cycles with flat hierarchy and strict RICE scoring, our task force achieved incredible numbers:

Out of 35 total tests, 8 delivered win-state value, 20 failed or stayed flat, and 7 were re-tested. But those 8 outliers didn’t just meet our initial goal ahead of time — they powered us past a stretched 6x target ahead of schedule.

From signals to scale: The path to personalisation

So, how do 35 tests, a handful of big wins, and a collection of flat results actually add up to testing your way to personalisation at scale? It comes down to a framework we rely on heavily at Zalando: Signals to value.

Every single experiment — whether a roaring success like stock scarcity or a localized mystery like our post-checkout “Woohoo!” — generates implicit and explicit customer signals. When aggregated across millions of shoppers, these signals uncover clear patterns of customer intent. Just like Sony discovered that a distinct subset of buyers genuinely wanted a yellow Walkman, we map these behaviors to specific customer cohorts.

By capturing live behavioral data and connecting those intent insights directly back to our RICE prioritisation loop, we can trigger targeted UX interventions rather than uniform changes. Personalisation at scale isn’t about guessing what 62 million people want from the start; it’s about building a continuous cycle that listens, learns, and delivers hyper-personalized value at speed.

The human impact: Keeping team motivation high

Beyond financial returns, the experimentation mindset transformed our team culture. As people managers, improving operational efficiency is only half the job. The real question is how this shift affects the team’s energy, morale, and professional growth.

“Beyond financial returns, the experimentation mindset transformed our team culture.”

In traditional product development, teams often ride an emotional rollercoaster. Spending a long time defining, designing, and aligning on a major initiative means delaying the impact. If the initiative underperforms at launch, the motivational impact is real.

Testing at scale completely changes that dynamic. When design, product and engineering see their work live in front of users within days, momentum stays consistently high. Scoping tests remove the stigma of failure — a flat result isn’t a defeat, but a valuable data point that prevents over-investing in the wrong direction. Most importantly, it empowers individual growth. When interviewing candidates at Zalando, we always ask what recent project they feel most proud of. Fast-paced experimentation gives creators tangible, data-backed success stories to own and take pride in. By giving our team the space to test, learn, and own their outcomes, we didn’t just smash a business target — we built a culture where innovation feels natural, rewarding, and fun.

How does your team handle the balance between long-term roadmaps and rapid experimentation? We’d love to hear how you test, learn, and keep your team’s creative energy high — drop your thoughts in the comments below!

Test your way to personalisation at scale: How we built a high-impact experimentation engine was originally published in Zalando Design on Medium, where people are continuing the conversation by highlighting and responding to this story.

The conversation around artificial intelligence in product design has reached a turning point. We’ve officially moved past the hype into an era where AI is deeply integrated into complex systems. As AI becomes part of our daily work, designers are facing real challenges, from rising anti-AI sentiment to flawed research synthesis. Navigating this shift requires more than just better prompts; it demands scalable design systems, resilient component architectures, and robust interaction guardrails.

To navigate this next phase, our design community gathered for Zalando Design Day, a full day event focused on the practical side of AI in design, cutting through the noise to share real-world insights. Featuring talks and workshops from industry expert Vitaly Friedman alongside our own designers, the event gave us actionable frameworks to use in our daily projects.

Read on for a deep dive into the practical strategies, interaction patterns, and mindset shifts that will help shape the future of product design.

Moving beyond AI hype to calibrated trust

When generative AI first exploded in 2022, the digital design landscape changed overnight. But as the initial novelty wears off, designers are encountering a very different reality. Kicking off the day with his keynote, “State of AI 2026,” Vitaly Friedman — founder of Smashing Magazine and Senior UX Lead at the European Parliament — explored how this shift is playing out on the ground. His primary takeaway? Designers are now facing a critical challenge: a growing “anti-AI” sentiment driven by confusion and a lack of trust.

“Designers are facing a critical challenge: a growing “anti-AI” sentiment driven by confusion and a lack of trust.”

Despite significant technical improvements, many users perceive AI as unreliable, especially when it comes to using it for work. Common pain points include increasingly subtle hallucinations, the heavy cognitive load of fixing AI errors, and a growing ‘AI fatigue’, backed up by research, where users often prefer non-AI labelled products, or products including “smart features” over explicitly labeled “AI features”. “AI-first” products are often perceived as being “cheap” or “untrustworthy” rather than innovative.

Building people’s trust: The designer’s new north star

To bridge this gap, Vitaly explained, the focus must shift from maximising trust to calibrated trust. This means designing interfaces that allow users to rely on AI when it is right, but easily verify and override it when it is wrong. Key strategies for building this trust include:

• Verifiable outputs: Instead of providing a “wall of text,” AI responses should include clear, navigable sources that point directly to specific segments of data.
• Keeping humans in the loop: Designers should avoid experiences providing full autonomy to agents. Instead, involve users even in small, approved steps to maintain a sense of control. Research shows that user resistance to AI labels disappears when features are framed as enhancing personal autonomy rather than taking away control.
• Context-aware interfaces: Moving beyond the blank chat box, the UI should change based on user intent — for example, shifting to a code editor when the user needs to write a script.

“Trust doesn’t emerge by default. It comes from understanding the product, interacting with it successfully, achieving goals — and then expanding usage. It’s all about consistent reliability.” — Shared Vitaly

Emerging design patterns for AI

The “2022 way” of manual prompt engineering is increasingly obsolete. Modern design patterns are evolving away from manual prompt engineering toward fluid multimodal inputs (voice, AI-enabled cursors), intent-based pre-prompts, and structured task builders.

Ultimately, the goal for designers in 2026 is not to replace human judgment but to determine when AI is truly valuable and when deterministic, reliable software is the better choice. By focusing on transparency and accountability, we can design AI experiences that people actually value and trust.

The paradigm shift: Designing interfaces for autonomous agents

In his session “Designing for no one”, Principal Product Designer Sriram Venugopal explored how the definition of a “user” is rapidly expanding to encompass both humans and autonomous AI agents acting on their behalf. We are entering the era of agentive design — crafting software that acts on our behalf rather than software we manually operate.

When the entity navigating your product isn’t human, traditional design playbooks break down:

• Agents have zero FOMO: Banners flashing “WOW Deal,” “Only 4 left in stock,” or urgency-driving countdowns have zero emotional impact on an agent. The agent executes the parameters requested and discards the visual noise.
• Agents do not reconsider: Agents don’t hesitate, they do what they are told to do. Safeguards and friction designed to slow down a human in their decision-making process do not work on an agent.
• Agents do not wander: They don’t leisurely browse lookbooks, stumble on things they didn’t know they wanted or explore casually. They need to be told up-front exactly what to look for; they don’t get distracted.

While we are not yet at the point where humans completely rely on agents to manage all of their daily tasks and purchases, more products are being built with this capability in mind. So, how do we design for an autonomous agent while keeping human trust at the core?

Designing for accessibility is the superpower for agent readiness

What an agent reads is the accessibility tree of your product. It values clean data, clear labels, and clear semantic structure over visual flair.

Internal testing reveals a stark trend: an agent’s task completion rate sits at 78% under standard operating conditions, but crashes to just 28% when structural accessibility is broken. The clean code, semantic labels, and accessible systems we build today are the exact frameworks AI agents will rely on tomorrow.

As designers, we play a crucial role in defining what an agent is allowed to do within our products. We know what the agent needs to navigate, as well as what the human user wants: to feel confident delegating tasks, to trust the agent’s actions, and to maintain control.

“We have two users with different needs. Designing for both is the work we need to do now” Sriram concluded.

Practical playbooks: Take these AI design insights further

Navigating this next era of AI requires both big-picture vision and day-to-day tactical shifts. If you want to explore the hands-on methodologies presented at Design Day, check out the dedicated playbooks published by our speakers:

• Why your generative customer insights are probably wrong (and how to fix them)
  Emilie Alonso, Principal Product Designer, pulled back the curtain on the hidden pitfalls of using generative AI for qualitative discovery. In her guide, she breaks down the 5 cognitive AI flaws in research synthesis and shares a system prompt framework to enforce data grounding.
• Beyond the prompt: mastering Figma Make for pixel-perfect design
  Yuls Walters, Product Designer, brought a wave of practical optimisation to the canvas with her talk, “Beyond the Prompt: Mastering Figma Make for Pixel-Perfect Design.” She introduces her “3Cs” framework (Content, Context, Correction) to help designers structure clean Auto Layout files, avoid prompt ambiguity, and escape generative loops.

Designing the next chapter of product design

AI isn’t here to replace us as designers; it is here to help scale, amplify, and empower the systems and structural thinking that already live within us. The real opportunity ahead lies in bridging the gap between human intuition and machine capability: defining ethical guardrails, building resilient systems, and creating intentional experiences that foster genuine trust.

How is your team navigating AI integration into your daily design process? Have you experimented with agentic UX or designing for both human and AI users? We’d love to hear how you’re approaching these shifts!

Exploring the future of product design with AI was originally published in Zalando Design on Medium, where people are continuing the conversation by highlighting and responding to this story.

Major innovation is happening across the board in design technology, especially with tooling built around large language models (LLMs). We’re finally moving past the old roadblocks, like getting bogged down by manual production work or feeling limited by coding skills. These new LLM-based applications let designers move more quickly from ideation to validation, with fewer bottlenecks.

The number of dedicated deployment tools that are already available — along with Figma’s native LLM integrations — definitely signal a permanent shift, particularly toward AI-assisted prototyping and MVP scaffolding. To find out exactly how new tooling might enable quicker, more interactive prototypes, I put it to the test as part of an improvement initiative on our product detail pages (PDPs).

Solving the PDP video gap

When shopping online, users need to see a product in motion and from multiple angles to truly assess it and make a confident purchasing decision. To test this hypothesis, we recently analyzed data from our PDP and found a clear solution: autoplaying a video as the primary asset in the image gallery significantly drives both conversion and customer engagement.

The challenge here is one of scale: there’s a substantial content gap across PDPs — a shortage of rich, video assets where customers make buying decisions. Many PDPs still feature static images only. To optimise the experience for these pages, we needed to rapidly design and validate high-impact interim solutions while video production ramps up. This challenge presented a perfect testing ground for using AI-enabled tooling to discover which approaches were the most feasible and impactful.

Three hybrid approaches to test

The goal of this experiment was not to ship a completed design, but to help stakeholders visualise the different possible solutions and speed up the decision-making process. We had three concepts that we wanted to test and evaluate:

Auto-advance image gallery

This approach simulates video by automatically cycling through the existing gallery of static images. Dynamic transition effects also help the page to feel less static and more experiential.

Slideshow video

This option compiles existing static images into a pre-rendered video asset with polished transitions between images. The video is then placed as the primary PDP media file. Although it can’t display a garment in motion, it showcases the product without adding friction to the user experience.

AI-animated images

Because we have extremely high-quality baseline photography across our PDPs, the third concept involved using generative AI to animate static images. The technical challenge here is that a single source image doesn’t show the item in multiple dimensions; the tool needs to synthesise motion without hallucinating product characteristics or misrepresenting the item. Testing this was important for determining if this approach would be viable.

“Design-optimised LLM tools can help me to put my ideas into the hands of my audience — in a way that’s simpler, more polished and more interactive.”

The results: simpler for stakeholders and easy to iterate

To test and visualise these three concepts, I used LLM-based tooling to put together a functional prototype for each one. I started with the production Zalando app wireframes as a baseline. Then I used targeted prompts to define the motion mechanics, including the transition logic and speeds. I also added a configuration panel within the interface so that I and other users can toggle and test different transition styles in real time.

Although the objective wasn’t to deliver a finalised end-to-end user journey, my work with the LLM produced a relatively sophisticated prototype. Not only was I able to validate the three different concepts, this approach let me adjust the prototypes directly, without going back into my usual design tool and rewiring everything.

Using this new tooling also had other practical, operational benefits. During the review meeting, stakeholders could open the live prototype links directly on their phones and experience the concepts hands-on as I was presenting. One definitive takeaway for me from this exercise is that design-optimised LLM tools can help me to put my ideas into the hands of my audience — in a way that’s simpler, more polished and more interactive.

A new skillset for designers

Bridging the gap between concept and realisation has always been a core challenge of design. Traditionally, creating high-fidelity prototypes required heavy engineering support early in the process. Now, whether using native Figma LLM integrations or plugins for external environments, smart prompting lets us quickly validate our ideas independently, allowing us to bring stronger, more tangible concepts to our product and engineering partners. However, mastering this new workflow also requires adopting a few key guardrails:

• Data governance: Excellent data hygiene is an absolute must when using these tools: Avoid uploading proprietary assets, protect internal concepts from public exposure, and use enterprise-compliant tools.
• Code degradation: LLMs are more effective for prototyping when you use them to execute a defined vision rather than for ideation or general exploration. Frequent, radical pivots create technical debt and unstable code fragments, so start with strict requirements or finalized sketches.

This approach to prototyping is catching on more widely, as more and more designers begin to experiment and familiarise themselves with the technology. Ultimately, my experiments with LLMs for prototyping have had clear results so far: this is a powerful new technical capability that expands a designer’s autonomy and helps teams to accelerate the path to validation.

I set out to test whether AI-enabled tooling could help me speed up my prototyping process, and it certainly did. But not only that, it let me step into a more active role as an end-to-end creator and strategist, and share my insights with other designers. When we cultivate this new skillset in working with LLMs, it empowers us as designers to broaden our focus and have a deeper influence on product development at every step of the way.

How are you using smart prompting to speed up your workflows? Share your experiences in the comments!

Keep reading

• How to build agentic CX frameworks: transitioning from static to agentic workflows
• User research synthesis or AI bedtime stories?
• The evolution towards agentic AI customer experiences

Accelerating design workflows with LLMs: a PDP prototyping case study was originally published in Zalando Design on Medium, where people are continuing the conversation by highlighting and responding to this story.

If you’ve ever opened Figma Make, typed a beautifully crafted prompt and got back something that was… let’s just say not quite what you had in mind, this is the article for you.

AI is incredibly capable — and incredibly literal. It takes ambiguity as permission to improvise. The moment you leave a gap in your instructions, the AI fills it with its best guess — which is rarely the same as your actual intent. That tension between the promise of AI and the reality of its output is what I’ve spent a lot of time thinking about.

The perfectionist confession

I’ve always been a bit obsessive about organisation. Meticulous layer naming. Clean file structures. Everything in its rightful place. As a child, my toys were sorted by colour and type — and that was entirely my doing, not my parents’. My colleagues used to tease me for it. Back when we were all designing in Photoshop and Illustrator, my hyper-structured layers were admired but also quietly resented. Creating and maintaining them was a bottleneck in my workflow, and my team simply didn’t care as much.

Here’s what I’ve since realised: that exact mindset — the one I used to think of as a personal quirk — is precisely what AI needs to thrive.

And that is where the 3Cs come in.

“AI is only as good as the content, the context and the correction you give it.”

The 3Cs framework

Working with Figma Make isn’t just about writing better prompts. It’s about building the right conditions before you write a single word. The 3Cs: Content, Context and Correction are the framework I use to do exactly that.

Think of it this way: Content is the foundation AI shouldn’t invent on its own. Context is how you communicate intent, not just instructions. And Correction is how you iterate surgically, not sweepingly. Together, they transform Make from an unpredictable tool into a reliable creative partner.

1. Content: build the foundation

Semantic layer naming is the map AI uses to navigate your design intent. When your layers are named Group 47 or Frame 338644172, the AI has no idea what it’s looking at. When they’re named card/product-image or button/primary/default, suddenly it can recognise, interpret and reuse your structural patterns with confidence.

Establishing a clear nomenclature is foundational. A good rule of thumb: if the layer name doesn’t immediately tell a developer what it represents, rename it before you open Make. You can even use Figma’s Rename layers with AI feature to do the heavy lifting here.

“You can’t prompt your way out of a messy file.”

Auto Layout: the grammar AI reads
Auto Layout can feel like a headache. I understand the temptation to skip it when you’re working quickly. But it’s fundamental if you want your generated layouts to hold together. Without it, every component Make generates risks breaking your grid. Think of Auto Layout not as a formatting choice, but as the grammar that AI reads to understand the relationships between your elements, especially if you want anything close to responsive behaviour.

Set up a sandbox file
Before you begin any Make session, create a dedicated Figma file purely for experimentation. Add just one or two fully polished, audited screens, linked to the design system, using the correct components and design tokens. This is your source of truth for Make. Iterate fast, break things and never pollute your production files.

2. Context: communicate intent before pixels appear

This is the biggest section of the framework and the most important. Spoiler: the quality of your output is almost entirely determined by what you put in before the AI generates anything.

Use external AI to build your prompts
You don’t need to start from scratch with every new Make session. You can use other AI tools — Gemini, ChatGPT or Claude — to help you build your prompts before you touch Make. They’re remarkably good at surfacing requirements and edge cases you might not have considered. Feed them your brief, your constraints and your design intent, and let them help you write a tighter, more specific prompt than you would’ve written alone.

Build a prompt cookbook
Once you have a prompt that works beautifully, save it. I keep mine in a Google Doc, categorised by use case. The goal is simple: never start from scratch. Over time, your cookbook becomes a personal library of proven approaches, and it compounds. Prompts that work for a product card layout can be adapted for a checkout flow. Patterns that solve one problem often generalise to ten others.

Establish anti-pattern rules
Tell the AI explicitly what NOT to do. Anti-pattern rules are underrated. Instructions like “never detach components”, “never override colour variables” and “never use inline styles” protect the integrity of your design system in a way that positive instructions alone cannot guarantee. Paste them into the instructions section of Make or at the very beginning of your prompt. Before you describe what you want, tell the AI what is off-limits.

Attach your design and your design system
Link your prepared Figma screens directly into your Make session. When the AI can see your actual design, it stops guessing and starts referencing. More importantly, add your primary design library directly to Make. This ensures the AI pulls exact production assets rather than generating generic clones. The difference in fidelity is immediately visible.

Start with “Do not code”
This is my favourite tip and the one I now use to open every single Make session. Before you describe what you want to build, type “Do not code.” Then tell the AI your intent and let it ask you questions. Resolve every ambiguity before a single pixel is moved or a single line of code is written.

I used this recently to lock down animation speed and easing curves before any output was generated. The conversation took five minutes. The alternative — correcting animation behaviour after the fact, compound change by compound change — would’ve taken far longer. Planning with the AI upfront is almost always faster than course-correcting after.

Curate a high-fidelity image gallery and brand fonts
At Zalando, images are critical to how we communicate quality. Early in my work with Make, I noticed that generic Unsplash placeholders were undermining the premium feel of everything I was prototyping. The fix was simple: build a curated gallery of brand-quality images and reference it in every session. You can host it in Google Drive or as a GitHub repository.

The same logic applies to typography. Zalando Sans is available as a Google Font, which makes our lives considerably easier. Paste the Google Fonts URL directly into your prompt or upload the web font file, and you will never see a font fallback in your prototypes again.

3. Correction: iterate surgically

The AI has done its work. Now it is time to do yours.

Work from form to behaviour
When refining a Make output, resist the urge to jump straight to pixel-level detail. My approach is to work from the general to the specific, in four stages:

1. Form: Get the layout and grid right.
2. Content: Make sure your components and assets are in the right places.
3. Design details: Check colours, spacing and typography.
4. Behaviour: Interactions, transitions and animations.

Trying to polish the details before the structure is solid is the single most common mistake I see. And it wastes significant time.

Know when to hit undo
Every correction you make touches the underlying code. The more you compound changes, the dirtier things get. Learn to recognise these three red flags — and stop the moment you see them.

• The loop: the AI suggests the exact same thing you just rejected.
• The domino effect: a fix resolves one issue and immediately breaks another.
• The argument: the AI says “let me try this instead” but nothing actually moves forward.

When any of these appear, stop. Hit undo. Rephrase from a different angle. Pushing through them almost always makes the problem worse.

What can you do tomorrow?

If you take one thing from this, let it be the 3Cs. But if you want to start putting them into practice tomorrow morning, here’s where to begin.

• On Content: Audit your layer naming and switch to semantic tokens today. Set up a sandbox file with one or two screens. Write five anti-pattern rules before you open Make.
• On Context: Use another AI to help you build your first prompt. Start your next Make session with “Do not code.” Save every prompt that works.
• On Correction: Follow the four-step refinement order — Form, Content, Details, Behaviour — and trust the undo button when you hit a loop.

You don’t need to overhaul everything at once. Start with the habit that feels most achievable and build from there.

As a designer, I used to think my need for structure was a quirk — something to manage rather than to leverage. Working with AI has completely reframed that. The very habits that once slowed me down are now the ones that make my AI output sharper and faster than most.

“You don’t need to be a perfectionist to master Figma Make. But you do need to embrace the beauty of order.”

Because ultimately, AI doesn’t replace us as designers. It scales the system within us.

Beyond the prompt: mastering Figma Make for pixel-perfect design was originally published in Zalando Design on Medium, where people are continuing the conversation by highlighting and responding to this story.

Designing a new typeface wasn’t originally part of the plan.

The past few years have been a key period of evolution for Zalando. The company has transformed far beyond online retail, expanding and innovating across different spheres. Visually, it was time to let the brand do some of the ‘talking’ about what the company means today.

This was the starting point for the Brand team at Zalando. Their initial idea was to update the visual identity, aligned with the company’s core brand attributes: bold, welcoming and fun.

Visually, it was time to let the brand do some of the ‘talking’ about what the company means today.

But with this kind of strategic clarity about how the brand should look and feel, bigger change was inevitable. These brand attributes quickly became the foundation for a much larger visual refresh and, eventually, for a whole new typeface: Zalando Sans.

Phase one: the wordmark

The classic Zalando wordmark was the first piece for the team to tackle. After years of building brand recognition and identity around the wordmark, the goal was to give it a refreshed look and feel — contemporary and modern. Zalando isn’t just an e-commerce or tech company; it’s also an entertainment, content and lifestyle brand. The typography of the wordmark needed to capture and communicate this.

To start, the Brand team looked at hundreds of different typeface styles and considered the range of connotations they have. There were quite a few directions on the table at first, and they deliberately expanded the design space to learn more about what was possible.

But fairly quickly, the team settled on something that looks and feels like a natural evolution. Having this updated wordmark made it clear what a strong impact fresh typography could have on the whole brand presence. The wordmark propelled the whole project forward. The next step: update all of the company’s typography with this new visual voice.

Phase two: the typeface

Because the wordmark typeface was a custom design, there wasn’t an out-of-the-box font that could deliver the same effect for a full range of typography needs. The team also wanted to change the approach to type pairing; the combination of a serif font for headlines and Helvetica for other text had some brand distinctiveness, but there was opportunity for more differentiation.

The updated wordmark made it clear what a strong impact fresh typography could have on the whole brand presence.

This led to a second phase of the project: partnering with brand agency Kurppa Hosk to design a completely custom typeface that matches the updated wordmark. A project like designing a custom typeface really benefits from specific typographic expertise. Beyond their work as brand consultants, Kurppa Hosk also has an integrated type foundry, which brought the specialised skills that the project needed.

The updated wordmark is a relatively restrained humanist sans font with some modern touches. Still, the team knew that the full custom typeface would need further design work so that it could communicate the uniqueness of the Zalando brand. It would also need to be optimised for legibility and come in different weights and widths.

By this point, it was also clear that the new typeface should be a core part of Zalando’s visual identity. That meant that it needed to have a lot of creative possibilities and be flexible enough to have a long life with the brand.

Generic humanist sans font → Zalando Sans

Once the first draft was done, the next step was to go through the typeface and find ways to inject character. It took several passes to get it right, but these adjustments ultimately made Zalando Sans come alive:

1. Wider cuts
2. Subtle curvature
3. Individual characters

In partnership with Kurppa Hosk, the Brand team spent time with characters like the K and the Q, shaping the curves so that they’re uniquely Zalando. The descenders on characters like the R and the Y also have stretching lines that support the typeface’s elongated and wider stance, without actually taking up more space.

Flexible enough for different propositions

Across Zalando, there’s a whole range of different needs and propositions. The new typeface needed to be expressive enough for all of these varied contexts.

As a creative brief, this is pretty demanding, especially when it came to text-based editorial content. In the past, the team had used a serif typeface more extensively for this kind of context. Without the serif, body text looked like it could be from any webpage — it didn’t feel inspiring enough. Paying homage to Zalando’s origins as a fashion destination, while still making room for new expressions, was another factor.

The team played with variants that have more of a fashion-heritage feel, without going into serif territory, but those didn’t end up being part of the core typeface. Instead, they experimented with typesetting. Rather than changing the variant, they were able to set the font in different ways to create a much better editorial feel for longer text pieces.

The feel of the typeface became breezier and a lot more interesting thanks to thoughtful typesetting choices: adjusting the leading and tracking, adding bigger text and pull quotes, increasing the use of white space and making use of the grid when structuring text.

Bold, welcoming, fun

All three of Zalando’s brand attributes were core to the type design.

Boldness doesn’t always have to be literally typeset in bold. Across the company, Zalando Sans Extended is the primary font variant, which is a wider setting of the type. It has a presence and a stance that absolutely equals boldness. (But yes, there’s a font variant in bold, too.)

A humanist sans typeface is an approachable choice; it communicates a point of view, but is welcoming to everyone. Zalando Sans is also very legible and easy to read, which makes it friendlier for people with different access needs. By designing a custom typeface, the team could make sure that it included all the necessary special characters for different languages — and there’s the opportunity to add new variants or characters at any time.

The fun is in the little, surprising details — such as, again, in the unique curvatures of some characters. But in a larger way, the font’s variability also makes it fun, because it’s a powerful animation tool. Making Zalando Sans literally move helps the font come alive.

Looking back — and forwards

The Brand team sprinted hard on this project, and it wasn’t a linear process. There were lots of twists and turns. All the underlying strategy work helped to keep things going; everyone felt clear conviction about why they wanted to update the brand, and there was alignment on the purpose and direction.

The typeface is just the beginning — that’s perhaps the most profound part. The team has already made tweaks here and there as they discover things that could work better for certain use cases, for example, branded merchandise or special documents like financial reports. Rather than having to licence a new typeface or do things by hand, they can expand what the font is able to do.

Usually, a design team just gets a font off the shelf, and it’s not possible to edit it — and some open-source fonts haven’t been touched in decades. But in the tech industry, fonts are increasingly treated as a software product. With Zalando’s foundations as a tech company, it’s a good home for a custom, open-source font.

Besides the practical business impacts — more efficiency, no licensing costs — a custom typeface is a compelling creative choice for real brand differentiation.

The future is wide open for Zalando Sans — literally. By putting it out open source, there’s also the option for folks from outside of the company to contribute and explore it. With the fresh, custom typeface, the Brand team had a chance to really say something new on behalf of the whole company — and to put the brand into practice. This bold step not only redefines the visual voice but also invites the wider design community to be part of its ongoing story.

Keep reading

• Bringing more character to the Zalando Design System with theming
• From the archive: Harnessing the engagement power of typography

The story behind the typeface: Zalando Sans was originally published in Zalando Design on Medium, where people are continuing the conversation by highlighting and responding to this story.

As designers, we tend to live in our own world. We reach for the tools, frameworks and solutions that feel familiar. It’s comfortable. But it comes with a risk of disconnection: how do we know if our work drives the results we’re aiming for?

The key to being effective business leaders is connecting our work to its impact. To strengthen this connection, designers need to cultivate their strategic skills and sharpen their business focus.

1. Speaking the same language

Aligning with business stakeholders starts with establishing a shared vocabulary. It sounds simple, but it’s a foundational challenge: across different products and platforms, even the most basic metrics can mean entirely different things. If we aren’t aligned on the definitions, we can’t be aligned on the strategy.

Take a metric as seemingly straightforward as a “view.” Inside the Zalando app, a view on the home screen, where a user interacts with a few large, curated content blocks, looks entirely different from a view on a fast-moving, high-density feed or a continuous video stream.

As designers, our job is to look past the surface of the UI and master these nuances. True collaboration happens when we intentionally bridge the gap between our craft and business reality, translating the pixels we build in Figma into the strategic outcomes tracked in Business Requirement Documents (BRDs). When we understand exactly how analytics and business teams measure success, we can design with much higher precision.

2. Designing the ‘why’

Design is fundamentally about exploring different solution paths, recommending a strategic direction and explicitly tying that choice back to business goals. This final step is what separates execution-focused designers from true strategic partners.

To lead effectively, we have to look beyond the immediate feature request and deeply understand the business mechanics that are driving it. This means knowing exactly what success looks like for the product, identifying the specific levers that move the needle and ensuring our design solutions directly influence those metrics.

Experimentation is central to our work, and designers are increasingly in the driver’s seat. While product and design teams collaborate closely on discovery, designers are now expected to take ownership of how their solutions move key performance indicators. We’re seeing designers take the lead and become critical drivers of product strategy and scalable business growth, from actively co-authoring experimentation hypotheses to analysing test results.

3. Following up on long-term impact

Designers are an active part of the execution of a project — there’s no question there. What’s sometimes missing, however, is the follow-up. We can look at our Zalando Loyalty programme to illustrate the value of long-term follow-up. We launched the Loyalty programme in our first markets in July 2024, but it took almost a year to start seeing definitive results. A proposition like Loyalty is built on customer retention, which makes it tricky to measure through short-term metrics alone. You don’t get instant gratification and instant results. Success takes patience and breadth.

This is why it’s vital for designers to stay close to what they’ve put out in the world. We need to stay deeply curious about how the work is performing, not just on launch day, but after three, six or even twelve months, evaluating both customer value and business impact. Setting the right foundation for testing and scaling –with this broader horizon in mind — is exactly how designers transition from tactical executors to long-term business leaders.

4. Upskilling and raising the bar

Our first Design Day for our Zalando Design Community was a deep dive into ‘Design talks business’. Beyond simply connecting our craft to its impact, the focus was on supporting our designers as business-minded leaders who co-author and drive the experimentation that fuels our growth.

The atmosphere was inspiring, filled with active questions from the audience and conversations that continued long after the panels ended. We encouraged our presenters to share through the lens of their own specific projects, offering concrete, tangible advice. An event like this is a major milestone for coming together and raising the knowledge floor of the entire design community, but it’s also a moment for individual self-reflection.

True upskilling is a process of putting theoretical concepts into actual practice. Since that first Design Day, we’ve seen our designers actively take ownership of their professional growth. They’re writing sharper hypotheses, tapping into the expertise of their peers, analysing past experimentation papers and finding ways to test and learn faster.

Growth like this doesn’t happen overnight. It requires designers to systematically identify gaps in their business skillset and proactively seek out opportunities to bridge them, whether that means adjusting their daily workflow or partnering with leadership to get exposure to new types of strategic work. Not only does this improve designers’ individual skills, it fundamentally raises the bar for what design can achieve for the business.

Closing the loop and staying close to strategy

Embedding a business mindset into the design process once might have been just an advantage, but now, it’s a definite necessity for meaningful innovation. Exceptional design work can’t be isolated from business reality. To elevate our impact, we need to go beyond solving user problems and intentionally close the gap between craft and strategy. This happens when we’re aligned on shared metrics, designing for clear strategic objectives, committed to long-term follow through and continuously upskilling our teams. In bringing this strategic mindset to our craft, we unlock sustained growth and centre design as a core driver of business success.

Design Day 2026: toward AI-driven innovation

Building on the momentum of last year’s inaugural event, we’re ramping up for AI Design Day 2026. This year’s full-day agenda is curated to help designers upskill and to equip them with practical, actionable takeaways for their daily work. From keynotes exploring the state of AI adoption and user research to lightning talks on cutting-edge creative tools, the conference is focused on uncovering relevant, real-world insights.

Looking ahead, we’re expanding on our core foundation: design as a strategic, business-minded function, while embracing AI as a critical tool for future innovation and growth.

Keep reading

• Beyond the keynote: step out of the audience and into the workshop
• Inside the strategic shift to implement our new Loyalty programme

Design talks business: four keys to unlocking impact was originally published in Zalando Design on Medium, where people are continuing the conversation by highlighting and responding to this story.

As AI agents become the new apps and operating systems, the very nature of digital experiences is changing, forcing a fundamental rethink of product design and development.

Product designers are no longer focusing on shaping interfaces and journeys. They’re increasingly shaping behavioural systems and creating experiences that can act on behalf of a user in a controlled and meaningful way. This does not reduce the role of design — it expands it.

To answer the burning questions facing design today, we need to move away from hype talk and chasing the latest AI design tool. There doesn’t exist a magical AI model that does the strategy setting, priority planning, and key design and development decisions for you. Instead, we need to move towards addressing the foundational needs and requirements of product design and development teams working in the age of agents.

Agents change the scope for product design and development

With emerging agentic capabilities, the nature of the digital experiences themselves is changing significantly. Product designers are not designing static interfaces and journeys, but rather, personalised autonomous or semi-autonomous experiences in which all the key elements of the customer experience can turn into changing and modifiable variables.

Concretely, user interface (UI), user experience patterns (UX), content (e.g. products, information etc.), data (e.g. customer behavioural data, product data etc.) and AI models (LLMs, classic recommendation models, etc.) turn into variables that together, as a dynamic combination, create the customer-facing experience. With agents, some of these CX variables can become invisible to humans, only relevant for other agents operating under the hood.

However, for agentic experiences, these are not the only variables that matter. To design systems that can proactively act on behalf of people, we also need to account for agentic features such as:

• State — what is happening right now in a task or journey
• Memory — what the agent retains over time about the customer, context and previous interactions
• Tools and actions — what the agent can actually do in the world: search, compare, and process information, create, recommend, transact, escalate or coordinate with other systems
• Goals — what the agent is optimising for
• Policies and guardrails — what the agent is allowed to do, when it needs approval, and when control must return back to the human
• Feedback loops and evaluation — how the system and the team learn what works and what doesn’t

In short, the CX variables can be tuned to create alternative versions of the experience using different combinations of variables and their values. As a simple example, to optimise your app experience, you might choose to change the UI module composition and UX pattern, but still use the same content, AI model version and customer behavioural data to power the experience. Or you might tweak the UX pattern and the AI model, e.g. through prompting, but leave UI and data as they are. Or, you might choose to simplify a part of the customer journey by distilling UI, UX, behavioural data, model and agentic features into a single pressable customer-facing prompt that then triggers an autonomous agent with a particular skill set, allowing it to interface directly with other agents without those interactions being visible to humans.

A framework for creating dynamic agentic experiences

The key design question in agentic experiences is no longer only: “what should the customer see, experience and interact with?” It’s also a question of what the system should know, remember, decide, do, optimise for, explain, defer and escalate.

To effectively design, develop and iterate agentic experiences, you need a framework and workflows that allow you to understand and modify the CX variables in a systematic, observable and controlled way based on your customer and business goals. Similarly, the workflows should allow you to create shared language across job disciplines and stakeholders, empowering everyone in the team — humans and agents alike — to build towards the same goals.

Such a framework and related workflows aren’t born automatically when you explore and experiment with different AI tools and tech stacks to build agentic experiences. Rather, you need to intentionally build towards them. So how might you get started?

Clarify your purpose and key use cases

What are you building towards and which use cases are critical for your customers and business? How does your solution and process create customer and business value? What kind of a setup enables the team and your organization to achieve its goals? Answering these questions forms a critical foundation that enables you to navigate the ambiguity in the fast-moving agentic CX and tech landscape.

At Zalando, we believe that agentic AI will transform the ecommerce business: the shopping experiences and the way they are designed and built. The agentic change spans from customer-facing experiences to internal tooling and logistics.

Purpose becomes even more critical in the age of agents because the system is no longer only presenting options but can increasingly make decisions, initiate actions and shape journeys on behalf of the customer. You need to be clear what kind of autonomy is desirable, where it creates value and where it creates unnecessary risk, friction or cost. In many cases, the best solution is not full autonomy, but carefully calibrated autonomy: the right balance between user control and proactive system initiative.

Define the core CX variables

Do you and your teams have a shared understanding of how different elements and underlying capabilities contribute to creating the experience? CX variables can be split into two main buckets: visible elements (UI/UX, content) and under-the-hood capabilities (data, models, agentic features of state, memory, tools, goals and policies).

Start by creating a shared understanding of how the visible and unseen variables come together in the current experience: this is UI/UX, these are the content, data and memory being used, these are the model(s) and tools that orchestrate the experience, this is the policy logic and these are the goals being optimised for. Creating such a shared understanding gives you a critical baseline to start assessing how the variables contribute to the experience and how modifying them might change it or build towards a future target state.

Importantly, these variables are not independent ‘knobs’. They form an interdependent system whose behaviour needs to be observed empirically. A change in memory logic, ranking model or approval threshold can have as much impact on the customer experience as a change in interface or content.

In agentic experiences, visibility itself becomes a design variable: what should remain visible and controllable for humans, what can be abstracted away and what must always remain visible and controllable.

Develop a workflow for modifying and iterating CX variables

The agentic era, with its transforming job descriptions and roles, requires us to rethink design and development workflows. To build functioning multidisciplinary agentic workflows with the right tools, product design needs to partner with other job disciplines, especially engineering and product, to build a concretely shared working flow and environment. This type of environment should allow product designers, engineers, data scientists, product and business people to collaboratively contribute to the evolution (and revolution) of the experience. Similarly, it should allow the team to fluidly and seamlessly bring and orchestrate agents in the workflows.

Imagine a tooling view of your experience that allows you to see how the different CX variables come together, enabling you to modify and compare the resulting CX in one view. The tooling should allow you to control and connect agents and tools (e.g. MCPs) that can then modify UI, UX, content, data, models and agentic features, informed by simulations and evals. It should also enable teams to inspect state, memory, goals, actions, approvals and outcomes so that the system behaviour becomes more accessible and observable across disciplines.

Any existing AI design or vibe coding tool doesn’t offer this kind of workflow off the shelf. The more dynamic the experience, the more it requires simulation and online/offline evaluation capabilities, since the variations and permutations of the variables can’t be captured by humans alone. You need to figure out through exploration, testing and iteration what works best for your team and organisation.

For agentic systems, such workflows also need to support decisions around autonomy and accountability. What can the system do without asking? Which actions require explicit approval? Which actions are reversible, and which are not? How should the system communicate confidence, uncertainty and next steps? From a workflow perspective: who owns the policies, prompts, models, memory and evaluation logic? The more capable the system becomes, the more important it is to make these decisions explicit instead of leaving them implicit between job functions.

Create a baseline and tooling for CX quality evaluation

The more dynamic and agentic the experience becomes, the less sufficient traditional design reviews, user tests and quality assurance methods become. You need evaluation capabilities that allow you to assess not only whether the experience looks and feels right, but whether it performs, behaves and creates the impact you’re after.

This means evaluating agentic experiences across multiple layers. At the most basic level, can the agent complete the task successfully? Beyond that, how helpful, proactive, coherent and trustworthy does the experience feel? How often do users need to intervene, correct or override? What kind of business impact does the experience create? What kinds of risks, failures or unintended outcomes emerge over time?

In practice, this requires a combination of qualitative research, high-fidelity prototyping, offline evals, simulations, online experiments and continuous monitoring in production. It also requires teams to agree on what “good” actually means across customer value, business value and system reliability.

Design for trust, failure and recovery

Agentic experiences are not only defined by how well they perform when everything works. They’re equally defined by how they behave when they are uncertain, wrong, interrupted or operating without necessary contextual knowledge. This requires teams to intentionally design beyond the ideal, ‘happy’ path.

How should an agent make its actions visible, or, for example, communicate uncertainty? When should it ask for clarification, and when should it proceed based on best judgement? How can a customer inspect, correct or reverse an action? What happens when memory is wrong, context is incomplete, the agent fails to use a tool or the model chooses the wrong next step? As systems become more proactive and semi-autonomous, managing failures gracefully through auditability, rollback and recovery become essential design topics, not just technical ones.

In evolving Zalando’s agentic tooling landscape, we’ve worked on foundational features related to agent observability, authentication, access management and related guardrails. Addressing such topics is pivotal in creating compliant and trustworthy agentic tooling at scale.

This is where human-centric design becomes even more important. Trust in agentic experiences is built not only through intelligence, but through usability, explainability and control. Customers and internal users alike need to understand what’s happening, why it’s happening and what they can do when they disagree with the system.

Be tool and tech agnostic to find the right tools and processes

Don’t chase the latest AI tool; focus instead on figuring out what tools you need to create value for your customer and business. There is no one AI tool or platform currently available that provides a way to take your idea from concept straight to market-ready and compliant digital product, and then iterate it forward by effectively combining and modifying the core CX variables.

Different AI tools come with different features and limitations. You need to figure out the tool stack, model environment, baseline agentic setup and data landscape that works for you. The requirements for choosing the right tools for the task depend on your organisation’s mission, working culture, existing systems and the company’s specific requirements, especially in relation to data, privacy, security and legal topics.

In Zalando Product Design, we’re working closely with our Engineering, Applied Science and Product partners in defining the requirements and finding the scalable tools for multidisciplinary workflows, bridging the gap between design and code.

At the same time, it’s important to recognise that not every use case benefits from agentic behaviour. In some cases, a simpler deterministic flow, transparent interface or classic recommendation system may create a better and more efficient experience. The goal is not to make every experience autonomous, but to identify where proactive and delegated behaviour creates disproportionate value compared to the complexity, risk and operational cost it introduces.

Success in agentic CX is mission driven

To drive product design in today’s agentic world requires us to rethink the foundational design inputs, outputs and workflows, and to embrace a cross-functional approach.

Not every AI-powered experience is truly agentic. Until it can interpret goals, select or plan actions, use tools, operate with some persistence across steps, and adapt its behaviour based on feedback or state (within bounded autonomy), a conversational interface or recommendation system alone does not yet make an agentic experience.

The agentic ways of working have the potential to super charge your organisation when they’re built around the mission, purpose and values of your team, not around the tools or technologies. And agentic experiences that succeed are not defined only by intelligence or automation, but by how well they solve concrete customer problems, balancing initiative with control, adaptability with explainability, and autonomy with accountability.

How to build agentic CX frameworks: transitioning from static to agentic workflows was originally published in Zalando Design on Medium, where people are continuing the conversation by highlighting and responding to this story.

Finding fashion inspiration used to mean flipping through Teen Vogue, copying looks from Sex and the City or (special shout-out to Gen X and elder millennials) switching on MTV. Today, we’re turning somewhere else to decide how to dress: social media.

People aren’t just using social media platforms for connection and entertainment. Brands have also populated our feeds with their products. Social media has changed how we decide what’s cool, how we develop what we like, how brands sell and how we shop.

Decoding the digital habits of Gen Z in Europe

When it comes to shopping online, Gen Z uses social media for everything. Their approach is less “shop ’til you drop” and more “infinite scroll”.

In Germany, almost all Gen Z uses social media and have used it to discover fashion items. Users often passively kickstart their shopping on platforms like TikTok, Instagram or Pinterest, and it’s not only attention-snagging paid ads that drive their purchasing choices. In fact, social media platforms have changed people’s basic expectations of content.

According to Zalando’s research, users often find TikTok content more personal and approachable because it features diverse, relatable, everyday people. For Gen Z, influencers are trusted guides. Whether through TikTok challenges, Instagram stories or YouTube reviews, recommendations from influencers feel like word-of-mouth style advice from a friend.

That affects the way users shop in Europe: 33% of Gen Z fashion shoppers look to influencers for inspiration, and nearly half have bought a product because of an influencer recommendation.

What users want

There are still traditional e-commerce platforms in the market that remain the place to shop for some users — but the social element is missing. Meanwhile, social media apps overflowing with inspiration. What’s worth calling out is that those apps are social media first, shopping destinations second: they’re being retrofitted for e-commerce, and that doesn’t always make for the smoothest purchasing experience.

Right now, if you spot something on Instagram or Pinterest you want to buy, it takes at least four steps to complete the purchase — or even more if you want to reverse search and look for similar products or compare prices on different platforms.

Based on our user research, there are four main issues that shoppers have with the current state of social commerce:

1. Lack of trust: With shopping on social media platforms, there’s often a lack of trust when it comes to payment methods and the legitimacy of service providers. Users tend to be nervous about logistics and payment information when products are being shipped from abroad.
2. Familiarity and ease of use (Jakob’s Law): The UX design principle of Jakob’s law says that users spend most of their time on other platforms, meaning they have an established mental model of how apps and websites should work. E-commerce has come a long way in the past couple of decades, so shoppers’ expectations are high.
3. Mismatch of expectations: A common issue with social media shopping is the mismatch between product quality and user expectation. Users often look for how some products look on ‘real people’ and not just models. Quality of material, proper size and fit are also big parts of decision-making when it comes to buying products online. They want to see the products in real-life environments.
4. Real-world context: When users are looking for inspiration, they’re also looking for different types of highly engaging fashion content like unboxing, ‘get ready with me’, style guides and edutainment videos.

Designing a social commerce experience

At Zalando, we’re driven to experiment, always looking for new ways to evolve the shopping experience. For me as a product designer, that often means asking how far we can push the boundaries of product innovation. When we’re designing solutions, we need to find the sweet spot between where user behaviour is currently at and where we believe it’s heading.

Based on our previous research into consumer habits, we already know that short-form video is a significant factor in how consumers discover products, seek advice and engage with brands outside Zalando.

To make social commerce easy and meet our users where they’re at, we’ve introduced videos with a ‘shoppability’ feature on the Zalando app. This lets people buy what they see instantly, blending shopping and scrolling seamlessly. It’s also available across all our markets in Europe, which lets us test how the new feature is working with users in different countries.

This feature is designed to tick the boxes of what users want and what they might find missing on other platforms:

1. Direct shoppability and safe payment methods: Our users know they can rely on Zalando as a safe place to shop — we built our platform to be secure, transparent, reliable and easy to use. This underlying infrastructure gives us an advantage with the ‘shoppability’ feature on our in-app videos.
2. Familiarity and ease of use: We know users are using other social media apps for inspiration and are used to short-form videos when it comes to discovering new fashion content. Therefore, we’ve introduced videos in the homefeed for users so they can find similar content based on their personal interests.
3. Trust and quality: When purchasing products on Zalando — whether it’s luxury, premium brands or mass market or sustainable brands — users are more confident about both product quality and brand authenticity. Our customers know they can expect that these products will offer an excellent fit and maintain high standards of quality.
4. Real-world context: Watching videos of people (who aren’t professional models) try on clothes helps users plan their outfits or visualise how an item might look on them. This preference for relatable and diverse body types is key to Zalando’s commitment to inclusivity. Seeing these videos might give users an extra assurance in the quality, colour and material of the product, which helps them in their decision-making and overall improves the shopping experience on Zalando.

Checking the facts: our follow-up findings

Since rolling out our new videostream and shoppability features in other parts of the app like Boards and Outfits, we’ve been conducting qualitative research and interviewing new and existing customers about their experience.

Our research confirmed that Gen Z users transitioned naturally to these features; the social commerce experience is familiar for them and gives them the context-rich information that they need when considering an online purchase. These updates also integrate smoothly with our other experience-led features, like Profiles and Stories.

New features aren’t just nice-to-haves. Over time, they redefine the mental model of e-commerce for millions of users. Moving from a transactional site to an inspiration-led social hub presents a massive playground for designers who enjoy solving complex, multi-demographic problems. We’re still in the early stages of personalising this video experience for everyone — and that’s where the real design challenge begins.

Is your For You Page your first stop for shopping inspiration? Are you a shoppability early adopter? Tell us in the comments!

Keep reading

• The 2026 shortlist: 11 design trends that are about to go mainstream
• Gen Z’s move to greater authenticity online
• Bringing IRL into digital with the Zalando Virtual Fitting Room

Scroll, see, shop: the one-click future of fashion was originally published in Zalando Design on Medium, where people are continuing the conversation by highlighting and responding to this story.

Design is moving at a lightning speed, and that makes clarity hard to come by. Cultural moments flare up and fade away overnight, while AI upgrades make yesterday’s ‘best practices’ obsolete before the week is out.

When systems are moving quickly, making predictions is easier said than done. We knew AI would be pivotal in 2025, and it’s been exciting to watch it truly take off. Even for savvy designers, mastering the underlying shifts of trends is where the real challenge — and the real surprise — lies.

Staying future-forward is more vital than ever. As we look at how the roles of designers and design leaders will evolve in 2026, we have to ask: what changes are already here, and what’s on the horizon? Eleven design leaders from the Zalando Design community lay out their forecasts for the year ahead.

1. AI will be agentic

Kalliopi Papadopoulou, Senior Product Designer

The “Help Bubble” is dead; long live agentic UX! In 2026, AI for e-commerce and retail will no longer sit on the sidelines — it will be driving core experiences, such as actively resolving logistics problems and guiding purchases. The primary design constraint is trust. We must build interfaces that verify intent, explain actions and know exactly when to step aside for a human.

2. Design work will transcend the technical

Yishi Yang, Principal Product Designer

The tools to build will finally be as fast as your thought process. AI will dissolve the technical ceiling, empowering designers to translate empathy directly into working software. Seize this capability to transform from a designer into a maker.

3. Designers will have more tools to explore

Oyindamola Akinleye, Senior Product Designer

AI advancements will further close the gap between idea and implementation, giving designers the power to turn concepts into reality with full control. This means we’ll see more designers extending the craft beyond static visuals. We’ll also see a rise in design-led companies and a wave of new, design-centric open source tools.

4. Design will be an experiment engine

Philip Ogaga Appah, Product Designer

2026 will not be about designing faster screens. It will be about designing smarter experiments. AI-integrated design systems will enable proactive exploration and rapid iteration, shifting our value to sharper problem framing and stronger hypotheses. Designers who master data and continuous testing will drive measurable outcomes, not just interfaces.

5. Design will be more cross-disciplinary

Sriram Venugopal, Senior Product Designer

Design specialisation will blur as AI handles more execution. The most valuable designers will be those who think strategically across disciplines — comfortable moving between product vision, engineering constraints and business context. Success will come from orchestrating experiences end-to-end, not mastering any single craft in isolation.

6. Design leaders will need to be hands-on

Inès Mir, Principal Product Designer

AI changes the most unexpected thing: how we see design leadership. We see how companies prefer hands-on individual contributor leadership to communicational design leadership. When leaders have the practical contributor skills that help them communicate faster and more efficiently, we can all get more work done that actually moves the needle.

7. Design will prioritise hands-on learning

Rafael Scovino, Senior Product Designer

Design’s future is action-led. Designers with a growth mindset will stay curious, experiment faster and adapt constantly. Those who can observe, reflect, learn and change quickly will shape what’s next.

8. True craft will be valuable

Çiğdem Demir, Senior Product Designer

While vibe coding democratises design and empowers everyone to build, the value of true craft becomes paramount. Those who have trained their soul, curiosity and eye will emerge as the most powerful drivers of change.

9. Design will go through its industrial revolution

Simone Poli, Principal Product Designer

Interface production becomes commoditised, but quality scales through process engineering. Like factory processes, designers will build AI ‘assembly lines’ that generate variants, enforce constraints and test outcomes across contexts. The job shifts from handcrafting screens to directing systems with intent, governance and evaluation.

10. Designers will need to prioritise intent over output

Andrea Brinkmeier, Principal Product Designer

When everyone can be a designer, AI brings a flood of algorithmic mediocrity in the disguise of smartness: superficial interfaces, meaningless text and images. Let’s make sure our discipline isn’t swept away, but elevated. AI can execute tasks while we master the heart. By offloading the ‘how,’ we can double down on the ‘why.’ Our job isn’t to compete with the algorithm — it’s to move the needle through empathy, expertise and an unbeatable sense of right and wrong.

11. Human creativity will be the differentiator

Thiago Hapner, Senior Principal Product Designer

In a world where everyone can build anything with the help of AI, human creativity will stand out and design decisions will need to be made based on true expertise, cultural knowledge and deep understanding of our customers.

The roadmap to design’s next chapter

The 2026 shortlist is now in, and these aren’t just temporary trends — they’ll become the new normal. Agentic AI at the forefront, leaders rolling up their sleeves, cross-disciplinary skills in top demand and a full-circle return back to authentic human craft: the design world’s moving quickly. Our forecast calls for a high chance of transformation — either way, the climate’s shifting.

What’s on your radar for 2026? Leave a comment and add your own predictions for where design is headed this year.

Keep reading

• How to go beyond the keynote and level up your learning
• Timeless design principles for the AI age
• The evolution towards agentic AI customer experiences

The 2026 shortlist: 11 design trends that are about to go mainstream was originally published in Zalando Design on Medium, where people are continuing the conversation by highlighting and responding to this story.

It’s the usual drill at the beginning of a new year — we set intentions for our personal and professional growth. For product designers, this journey often begins with a familiar path: attending conferences, bookmarking talks and mentally logging compelling arguments for change.

From my experience in DesignOps, I know that inspiration alone won’t close the gap between ambition and execution. Skills like code-driven prototyping or speaking the language of commercial strategy are critical for the design industry. But designers can’t passively absorb these capabilities — they need to be actively developed. And that’s where DesignOps needs to step in, to help create the low-stakes, high-impact learning environment where this can happen.

Dodging the passive learning trap

In a discipline as iterative as design, relying solely on lectures and keynote addresses can leave us with a mountain of notes and takeaways that don’t have much practical application. The skills that truly drive influence — strategic thinking, business literacy, cross-disciplinary building — require designers to step out of the audience and into the workshop.

High-quality, focused and hands-on learning experiences are rare. Knowing that, when the Design Matters team reached out with the wonderful suggestion of hosting a masterclass series in our Zalando office, our DesignOps team immediately confirmed. It was a unique opportunity to provide the kind of intensive, practical learning our design community needed to tackle two major areas of professional evolution.

We designed the Design Matters Masterclass as a one-day immersion focusing entirely on actionable skills and real practice. Together with Design Matters, our DesignOps team hosted over 60 designers across two parallel tracks, with intensive collaboration in small, focused groups.

The format worked, according to participants like Kim Does, Principal Product Designer: “I liked the tables with smaller groups — it was a good chance to talk with people, [and] it was good when they made us switch groups.”

The success of this approach hinges on the balance between theory, live demonstrations and immediate individual or group practice. For instance, in our AI-focused sessions, specific whiteboard templates allowed everyone to “get their hands dirty” — as Göksu Kaçaroğlu, Principal Product Designer described it — immediately rather than just watching a demo. A conversational approach where participants build on a speaker’s points through group discussion also made learning feel like a “real class,” Göksu said, adding: “The conversations among the masterclass participants, stimulated by the hosts, were very insightful and made the learning experience feel more practical.”

Prioritising the practical

We structured the Design Matters Masterclass around two parallel tracks, truly focused on doing and claiming strategic influence.

The first track, ‘Designer, Claim Your Future Seat’, concentrated on essential leadership and commercial accountability.

In his workshop, Will Shaw offered a framework to connect creative work directly to business goals like revenue and cost control. Next, Ram Yoga equipped designers with the math and quantification tools to ground their impact. The final session, led by Alex Simon, helped designers step into an innovation mindset and turn future trends and emerging technologies into practical, tangible ideas.

The second track, ‘Designer, Activate Your Future Skills’, tackled the immediate, hybrid reality of AI, automation and cross-disciplinary work.

One hands-on session with Oyindamola Akinleye and Felipe Santana focused on how to build with AI to automate repetitive tasks. Another workshop with Will Shaw on integrating AI into the designer’s toolkit while also designing for AI as a new distribution channel rounded out this track. The final session, led by Jean-Baptiste Kaloya, used the Design Leadership Archetypes framework to identify and amplify strategic style.

For me and the rest of our DesignOps team, it was key to have this dual focus on strategic influence and technical mastery to get participants curious and give them immediate skills to take back to their work.

The results: renewed energy to keep going and growing

Even though hands-on learning takes more focus and engagement, it’s also more generative. Madeline Rogier, Senior Product Designer, said that this approach made her want to go deeper: “The last workshop about design archetypes stood out for me the most. It got me curious and I even went home afterwards to continue learning more. I felt energized after the event, which doesn’t always happen after a long conference event.” Not only that, but the interactive format fostered deep, organic connection, she said: “The event made it easy to meet and talk to people, I even connected with several new people on LinkedIn.”

Patricia Evangelista, Senior Product Designer, pointed out that the workshops “taught us a lot of skills regarding ‘business first’ when it comes to justifying my work, and the materials will be super useful for articulating ideas with a solid business rationale.”

For Kim, the masterclass was also a confidence boost for the future: “Fresh ideas? Partially, also some confirmation that I’m not too bad at what I do, but also a nice spark of things I could be doing.” This moment of confirmation for a designer — that their talent is solid and they know how to identify their next growth frontier — is a cornerstone of professional development. It validates how much experience they already have, while motivating them to keep experimenting and innovating. This is exactly the kind of individual and community growth that DesignOps can — and should! — help to cultivate.

Practical tips for taking your team beyond the keynote

Based on my experience and the feedback from participants, here’s how DesignOps can make internal learning more impactful for our communities:

• Mix and match: Networking shouldn’t be forced. Create organic opportunities for designers to work with non-designers and peers from different contexts or countries. This diversity allows participants to see how others interpret and apply the same tools differently.
• Prioritise business know-how: There’s a massive appetite for designers to level up their business focus. Provide resources on how to talk metrics, justify designs with numbers and speak the language of product leadership. Materials like printed resource sheets on important business metrics are often more valuable than a high-level talk.
• Manage energy levels: Hands-on learning takes effort. Structure the day so that heavy, technical content (like AI building) is scheduled when energy is high, and balance that out with conversational sessions or lighter, creative workshops.
• Make it flexible: Finding the right size and fit applies to learning, too. In the future, we aim to offer more ‘choose your own adventure’ schedules where participants can combine individual classes from different tracks.
• Create solid, shareable assets : If you can’t record the sessions for team upskilling ‘watch parties’, make sure to share the slide decks and frameworks (like prompting templates or context-building guides) as widely as possible.

As DesignOps, we’re building the infrastructure for our community to succeed. Passive consumption might check the ‘professional development’ box in the short term, but it doesn’t enable our designers to reach their highest potential. Engaging, hands-on learning helps turn abstract inspiration into core competencies — and that’s what shapes designers into future-proof leaders.

Beyond the keynote: step out of the audience and into the workshop was originally published in Zalando Design on Medium, where people are continuing the conversation by highlighting and responding to this story.