Zalando didn't debate AI's creative merit. They understood their workflows, found bottlenecks, and compressed them.

Somewhere in your business, there's a process that takes three weeks when it could take three days. And because nobody's really dug into it and understood it, you're losing time and margin whilst arguing about whether AI understands aesthetics.

Zalando cut editorial content production from six to eight weeks down to three to four days. Pandora unified fragmented merchandising systems creating multi-week delays through o9 Solutions Digital Brain platform. These weren't creative breakthroughs. They were operational ones.

Three Candidate Workflows to Examine

Time Wasters (What's Getting Automated)

Pure data processing taking weeks for no defensible reason. Zalando compressed editorial cycles through AI-assisted image generation, with 70% of campaign assets AI-generated. Pandora consolidated demand planning, assortment, and financial planning into unified platforms. Leading brands replaced manual supplier coordination with real-time cost tracking. These are multi-week waits caused by manual data transfer.

Judgment Layers (What Stays Human)

Strategic decisions requiring market intuition. Zegna's AI consolidates customer data and suggests styling, but sales associates finalise every interaction. LVMH's data science team supports brands, but creative directors retain final authority on assortment. AI detects microtrends; merchandisers make the brand-fit decision. These are decisions where judgment, not speed, creates competitive advantage.

Hidden Margin Killers (What Nobody's Measuring)

Workflows draining margin silently. Fashion's days inventory outstanding ties up cash in overstock. Legacy systems require manual data transfer between departments, creating two to three week delays in processes that could be real-time. Brands still review pricing quarterly whilst consumers use real-time comparison tools. Forty-seven per cent of retail employees want AI training; nearly half report inadequate support. Ninety per cent of AI initiatives fail to scale beyond pilot phase. These are operational inefficiencies hiding because nobody's looking hard enough.

What You Could Do Now

1. Map your current workflow

Pick one high-volume process: product listing, inventory updates, content creation. Track when it starts, where it waits, when it finishes. Map every handoff.

2. Identify human judgment versus data processing

Does this step need human judgment (strategy, taste, relationships)? Is it pure data processing (reformatting, copying, verifying)? Is it waiting for approvals?

3. Calculate your "processing tax"

If your three-week workflow could be three days, what's the margin cost of those 18 extra days? Multiply by volume.

Our time is precious and it costs

A lot of businesses can't tell you how long it takes to get a product from photography to live listing. Or how many days inventory sits before conversion. Or how many hours merchandisers spend reformatting data between systems.

The question isn't "what should we automate?" It's "what's taking three weeks that could take three days if we actually looked?"

Pick one workflow. Time it properly. Then tell me it couldn't be faster.