Case Study
+15 % “take‑rate” lift (core revenue metric: items kept per shipment)
‑30 % cycle‑time to deliver curated products to customers
Model deployment time: 6 months → < 1 week
Company‑wide AI adoption across marketing, product, sales & ops
Executive Snapshot
- Client – Trunk Club (Nordstrom’s personalized‑fashion subsidiary)
- Strategic Challenge – Predictive models and data sat in silos; collaboration was minimal, reporting slow, and AI’s value invisible to most of the business.
- Bennett Data Science Solution – Embedded data scientists on every major team, rebuilt the data warehouse (30 tables → 3), stood up CI/CD for models, and launched real‑time AI visualizations.
- Enterprise Impact
- +15 % “take‑rate” lift (core revenue metric: items kept per shipment)
- ‑30 % cycle‑time to deliver curated products to customers
- Model deployment time: 6 months → < 1 week
- Company‑wide AI adoption across marketing, product, sales & ops
Client Context
Trunk Club delivers curated apparel boxes (“trunks”) to style‑conscious shoppers. As competition heated up, Nordstrom leadership mandated: “Make AI tangible, fast—and tie it to revenue.”
The Strategic Challenge
Bottleneck
Business Cost
- AI in a silo—models existed but rarely touched the customer journey
- Months before new hires saw model impact; wasted investment
- Bloated warehouse—400‑line SQL chains for a single report
- Analysts spent days on maintenance instead of innovation
- 30 overlapping data tables
- Slow model development; high error risk
- No collaboration channels
- Marketing & product unaware of AI capabilities
Our Approach
1. Embedding AI Into the Business
- Placed a data scientist on every core team—marketing, product, sales, ops
- Created direct feedback loops; questions answered in days, not quarters
2. Modernizing Data Infrastructure
- Re‑architected warehouse from 30 to 3 canonical tables; automated ETL
- Simplified SQL; enabled self‑service analytics and faster experimentation
3. Speeding Model Delivery
- Built internal CI/CD pipeline—train, test, deploy in < 7 days
- Allowed rapid iteration aligned to sprint cycles
4. Making AI Visible & Actionable
- Launched real‑time dashboards showcasing model outputs & impact
- Transparency built trust, sparked new AI initiatives organically

Quantified Results
”Bennett Data Science turned AI from a back‑office science project into a growth engine. We now innovate in weeks and see immediate impact on customers and revenue.
VP Product & DataTrunk Club (Nordstrom)
Why It Worked
Success Factor
- Embedded analysts—AI lives where decisions happen
Executive Value
- Faster iteration; deep business alignment
- Lean, governed data model
- Reliable insights; lower maintenance cost
- CI/CD for models
- Agility to meet market changes in real time
- Transparent dashboards
- Trust & collaboration across the org
Partner with Us
Ready to turn operational drag into bottom‑line growth?
Let’s discuss how Bennett Data Science can deliver a 3‑5× ROI for your organization.