Wehkamp: how onboarding 500 suppliers stopped being manual work

Company: Wehkamp, one of the largest online retailers in the Netherlands. My role: product strategy and delivery, three AI-powered B2B products. Worked with C-level stakeholders and engineering teams, 40+ microservices.

Challenge 1: suppliers were onboarded by hand

Every new supplier meant manual work: matching data formats, sorting products into categories, checking attributes, loading everything into the database. At Wehkamp’s scale this was the bottleneck holding back catalog growth.

What we did

Rules first. Simple conditional logic covering the most common cases: obvious format matches, standard categories, standard attributes. That delivered a quick win and, more importantly, a labeled dataset.

Then predictive models on top of that data: validation and categorization without hard-coded rules, wired directly into the database.

Result

• −40% time and effort spent onboarding products
• +30% data accuracy
• 500+ suppliers in the system

Challenge 2: stock and margin

Prices changed rarely and by hand, so the warehouse accumulated stock that wasn’t moving, and discounts went out at random and too late.

What we did

A machine learning model that reads demand patterns and stock levels and adjusts price in real time — so turnover goes up and warehousing costs go down.

Result after six months

• −30% excess stock
• +15% margin

Also

Scaled P2P payments with Mastercard and Paysend integrations, growing transaction volume. Plus in-person negotiations and workshops: some launch bottlenecks only get solved face to face.

What this means for your store

The scale is different, the mechanics are the same. If you or your manager spend every month manually onboarding products from suppliers, reconciling price lists in Excel, and marking things down by gut feel — you are doing exactly what Wehkamp did before these projects, just without their headcount budget.

For a store on WooCommerce or Shopify that looks like: automatic price-list ingestion, generated descriptions and attributes across the whole catalog, markdown rules driven by turnover instead of intuition. Things like this pay for themselves faster than any ad spend, because they cut the cost of your time, not the cost of acquiring a customer.