Zalando's One-Year AI Leap: Content From Near Zero to 90 Percent, Campaigns From Six Weeks to Days

Zalando SE, Europe's largest online fashion and lifestyle platform, scaled AI-generated marketing and product content from near zero to 90 percent of its total output in a single year, compressed campaign creation from six weeks to a matter of days, and lifted content production by 70 percent. The Berlin-based company disclosed the figures in its official full-year 2025 results on March 12, 2026 — the same document that builds its 2026 guidance of 660 to 740 million euros in adjusted EBIT explicitly on scaling AI across the business.

The situation before: six-week campaign cycles in a market that moves in hours

Fashion e-commerce rewards speed above almost everything else. A trend surfaces on social media, peaks within days, and fades while traditional marketing pipelines are booking photographers. Before 2025, a typical Zalando campaign needed six to eight weeks from concept to publication: studio shoots to schedule, agencies to brief, imagery to localize for a base of 62 million active customers across 25 European markets. Each campaign variant multiplied cost and coordination. By the time content reached the storefront, the cultural moment that inspired it had frequently passed. For a platform whose entire proposition rests on being where fashion happens first, that lag was a structural handicap — and volume constraints left merchandising teams with far less content than they needed. AI entered the picture as the simultaneous answer to both problems: speed to trend and scale of output.

The decision that made it possible: an in-house AI pipeline, owned end to end

Zalando's pivotal choice was organizational before it was technological: the company moved content production in-house onto a generative AI pipeline, in place of incremental agency augmentation. The architecture, described by VP of Content Solutions Matthias Haase in the company's own technical account, "How Zalando tells better stories", runs as a loop: real-time trend detection drawn from the behavior of more than 50 million customers on the platform, an internal social-listening tool that validates whether a signal carries broader cultural momentum, AI systems that generate initial visualizations, and creative teams that refine the output — deliberately preserving what Haase calls the "messy, human details" and authentic lighting that separate credible fashion imagery from robotic perfection. Humans keep editorial control and define, in the company's words, taste and the final vibe. By early 2026 the pipeline had produced 170,000 AI-generated product videos, and the company now targets trend-to-publication cycles of under 24 hours. The in-house choice mattered because it turned content into an engineering discipline: measurable, iterable, and compounding with every campaign, the same way Zalando already treated logistics and recommendation systems.

The result — with full context

The numbers Zalando placed in its results release are unusually specific for AI disclosure. AI-generated content went from near zero to 90 percent of marketing and product content in one year. Campaign creation dropped from six weeks to days. Total content output rose 70 percent, and customer engagement with that content increased 10 percent. The content pipeline sits inside a broader AI program with equally concrete results: delivery-promise precision improved by 22 percentage points, a technology organization of roughly 3,000 engineers shipped over 20 percent more code changes, the Zalando Assistant reached 6 million users — a fourfold increase — the Size & Fit feature avoided more than 8 percent of size-related returns, and foundational-model matchmaking lifted items added to bags by 13 percent. The financial frame: group revenue of 12.3 billion euros, up 16.8 percent, and adjusted EBIT of 591 million euros, up 15.6 percent.

Honest context matters here. As WWD reported, the About You acquisition drove a large share of headline growth; Zalando's core retail business grew organically in the mid-single digits. The efficiency drive also carries a human cost: the company is closing its Erfurt logistics center by September 2026, a decision affecting around 2,700 jobs that has sparked political controversy in Thuringia. And a 90 percent AI share means traditional photography workflows have been largely displaced; Zalando frames the change as creatives moving toward direction and curation, while the disclosure stays silent on net creative headcount. Co-CEO Robert Gentz's claim that AI delivers "experiences and services that seemed impossible just a few years ago" now reads as a forward commitment as much as a retrospective: the 660-740 million euro EBIT guidance makes 2026 the year the market tests whether AI-driven acceleration continues.

What other organizations can learn

The most transferable lesson is evidential before it is technical. Zalando put its AI metrics inside an earnings release — a document subject to securities law, auditor scrutiny, and analyst cross-examination — rather than a marketing blog. That choice imposes discipline upstream: every figure needs a baseline, a measurement method, and a time horizon, because investors will ask. Organizations aiming to replicate the result need four conditions. First, a proprietary demand signal at scale: Zalando's trend detection works because 50-plus million customers feed it daily. Second, ownership of the pipeline: renting creativity from agencies leaves the compounding gains outside the company. Third, human editorial control as the quality gate: the 10 percent engagement lift suggests customers respond to AI content when humans curate it. Fourth, baseline measurement before deployment, so before-and-after claims survive audit. The pattern generalizes well beyond fashion: any content-intensive business holding rich first-party demand data can compress its production cycle the same way. The bar Zalando has set is the disclosure standard itself — AI results reported with the same rigor as revenue.