How Hunkemöller Built an AI-Ready Data Platform and Team with Rittman Analytics

The Download

Hunkemöller is one of Europe’s leading lingerie brands. It was founded in Amsterdam in 1886 and today operates over 700 stores across 20 countries, alongside its own e-commerce sites, franchise partners and a large loyalty programme.

Every part of that model runs on data:

• Store teams start each day with yesterday’s sales, conversion and productivity against target.
• Regional and head office teams manage performance across the estate.
• E-commerce now accounts for a large and growing share of revenue, so the digital team needs a clear view of how customers move from browsing to buying.
• The run from Black Friday to Christmas is the most important trading period of the year, so the numbers have to be right when they matter most.

The Challenge

Over three and a half years, Hunkemöller moved from SAP transactional reports and team-by-team Excel to a modern data platform on Google Cloud. It built a BigQuery data lake and rolled Looker Enterprise dashboards out to the point-of-sale systems and iPads in its stores, replacing daily PDF reports. Adoption followed, and Looker went from around 200 users to around 6,000.

That success brought a new set of demands.

• Scale changed the economics: Most new users were store managers checking a small set of fixed KPIs, which should have been cheap to serve. Instead, Looker and BigQuery costs grew faster than usage, because the platform had been designed for a much smaller audience.
• Store teams needed simpler, faster dashboards: Retail leadership wanted dashboards built around four or five key metrics, refreshed reliably every morning. Combining data held at different levels of detail made filtering and refresh behaviour inconsistent.
• E-commerce needed clearer insight: After a recent move to Snowplow, the e-commerce team needed a dependable view of web performance and the customer journey.
• The business needed help asking the right questions: Turning requests into dashboards that answered the real business question needed analytics experience as well as development capacity.
• A growing team needed a common way of working: Hunkemöller was expanding its data team and focusing it on delivering insight to the business. That required one modelling standard, one delivery process and one place for business rules, so new engineers and partners could work consistently.
• Foundations for what comes next: Dwan’s longer-term aims were to raise data literacy, open up self-service analysis and prepare the platform for conversational and AI-assisted analytics. Each of these depends on a well-modelled, well-described semantic layer.

Hunkemöller’s own team and its warehouse partner had strong data engineering depth. Dwan wanted to add specialist experience in Looker Enterprise and semantic layer design, from a partner who could assess the setup independently rather than simply add development capacity. He had followed Rittman Analytics’ blog for several years and had already implemented its approach to RFM analysis inside Hunkemöller.

The first deliverables the business was waiting on were web performance insights and simplified store dashboards. Further investment depended on them.

The Solution

Rittman Analytics followed its three-step data team enablement approach. Over just over a year, it moved from a two-week diagnostic to a new analytics platform, a set of production dashboards and a Hunkemöller team that now owns and extends the work itself.

Step 1: Confronting Hard Truths. The Diagnostic Deep Dive

Rittman Analytics spent two weeks inside Hunkemöller’s data function in July 2025, interviewing retail, e-commerce, buying and data stakeholders. We reviewed their Looker instance, dbt repositories and reports currently in-use and mapped the set of KPIs most requested by users against those already provided by their platform, and discovered that:

• Foundations were stronger than they looked: Most of the data the business needed already existed in the warehouse. The gap was in how it reached users, not in the data itself.
• Logic was however in the wrong place: Complex business rules sat in Looker, where they were hard to test, slow to run and expensive at scale. Moving them into dbt and BigQuery, consistently across every project, would fix performance and cost at the source.
• The team needed a shared way of working: Requirements arrived without a clear business question, and each piece of work followed its own path. A single product development lifecycle, from idea to release, would make delivery predictable and reduce rework.
• Start small, prove value: Rather than wait for a complete warehouse, the recommendation was to deliver two priority use cases first, store performance and web performance, and use them to establish the new method.

Hunkemöller adopted these recommendations as their team’s strategy, and shifted its hiring towards dbt-focused analytics engineers as a result. Our diagnostics also produced a gap analysis, entity roadmap and interactive dashboard prototypes so that stakeholders could see and approve what they would get before any engineering started.

Step 2: Creating the Change. Architecting a Scalable Foundation

We worked alongside Hunkemöller’s data team to make reporting more reliable, with clear definitions that teams could use consistently. We also improved how changes were developed, checked and documented, making the platform easier to maintain and build on.

• Shaping dashboards with business users: We used interactive prototypes to agree what each dashboard needed to show. We then built it in stages so users could check the figures and give feedback along the way.
• Supporting everyday decisions: Together, we delivered dashboards for retail performance, store productivity, sales incentives, product readiness, e-commerce and campaign reporting, helping teams find the information they needed.
• Making reporting easier to use and maintain: We enabled teams across stores, regions and countries to use shared dashboards, with each person seeing the information relevant to their role. We also made dashboards faster and easier to update as needs changed.
• Keeping web reporting consistent: As Hunkemöller’s web analytics developed, we brought historical and new data together so teams could continue tracking performance and understanding longer-term trends.

Step 3: Managing the Change. Embedding a Culture of Ownership

From the start, our aim was to give Hunkemöller’s team the skills and confidence to run the platform independently.

• Learning through hands-on work: Their new analytics engineers worked alongside us on real projects. Within months, they were building data models and dashboards themselves.
• A clear, shared way of working: Together, we documented how to develop, review and complete work, including a step-by-step guide to adding new data models. We regularly reviewed and improved these guides as a team.
• Helping analysts build their own reporting: We enabled business analysts to create reports using trusted data, with checks and peer review to maintain quality. We also trained retail analysts to explore the data and answer their own questions.
• Introducing practical uses of AI: We showed the team how AI could help turn written requirements into dashboard prototypes and support development. We first tested the approach on a campaign dashboard, then made it reusable and taught it through training sessions and regular drop-in support. A store-level AI Coach is now being piloted.
• Preparing the team to take ownership: We walked the team through the platform, its tools and ongoing maintenance needs. After working alongside us during the handover, they took full ownership on 1 September 2026, with further handovers for individual dashboards and areas of work.

Business Benefits Delivered

The work gave Hunkemöller more reliable reporting, a simpler platform to maintain and a team equipped to develop it independently.

• Reliable reporting for stores: The Retail Performance dashboard went live in May 2026, bringing several older reports together. Store Productivity followed for regional managers and store teams, giving them regularly updated information to support daily decisions.
• Less duplication and lower running costs: Shared dashboards replaced separate reports for different audiences. We also identified older reports and duplicate data that could be retired, reducing maintenance and running costs.
• A simpler set of reporting tools: We moved retail productivity and campaign reporting onto Looker, supporting Hunkemöller’s move away from older tools.
• Consistent figures across dashboards: Dashboards now use the same trusted data and definitions. This also makes new reporting quicker to build, as teams can reuse work already in place.
• Quicker development with early feedback: Interactive prototypes helped users shape dashboards before development began. AI-assisted development helped the team turn agreed requirements into working reports more quickly.
• A team ready to take ownership: Hunkemöller’s engineers took full ownership of the platform in September 2026. They now plan and deliver their own work using shared, documented practices. Their retail analytics lead described the team as “much further ahead”, with “two very high-impact dashboards and more to come.”