Client Perspectives
A selection of perspectives from Singapore retail and e-commerce businesses that have used our personalisation, planning, and segmentation services.
Back to HomeClient Reviews
Charmaine Lim
Head of Digital, Fashion Retailer — Singapore
We had been talking about personalisation for two years without making any real progress. Shelvik came in, looked at what we actually had, and scoped something practical rather than aspirational. The documentation they left behind is something our developers still refer to. It wasn't the biggest project we ran last year, but it was probably the most useful.
February 2025 — Personalisation Engine
Raj Thiruchelvam
Buying Manager — Electronics, Singapore
The demand planning model has changed how we prepare for the year-end and Chinese New Year peaks. Before, we were relying on gut feel and last year's numbers. Now we have a model that incorporates the promotional calendar and I can actually explain to my team why the numbers look the way they do. Shelvik were clear about what the model could and couldn't do, which I found refreshing.
January 2025 — Demand Planning Model
Michelle Hartono
Founder — Home & Living E-Commerce
I started with the segmentation analysis because SGD 680 felt like a reasonable amount to spend to find out whether there was anything interesting in our data. There was. We had one segment we hadn't properly identified — high-frequency, low-average-order buyers who responded very differently to email than our other segments. That insight alone changed our next campaign. We've since done the personalisation project with them.
December 2024 — Segmentation Analysis
Yusri Karim
Operations Director — Multi-Brand Retailer
What I appreciated was that they were honest at the beginning about what our data could support. We had hoped to do more aggressive personalisation but Shelvik told us our browsing data wasn't clean enough yet and walked us through what we'd need to fix. We addressed that over three months and then came back for the full project. That kind of transparency is rare.
January 2025 — Personalisation Engine
Sandra Tan
Head of Marketing — Wellness Retail
The segmentation narrative they produced was something I could share directly with our agency. Not a slide full of cluster charts — an actual written account of who our customers are and what drives their behaviour. My team found it useful in a way that previous analytical outputs hadn't been. The framing made a real difference.
February 2025 — Segmentation Analysis
Daniel Ng
CEO — Sports & Outdoor, Singapore
We've tried two other AI vendors for demand planning and neither stuck — the outputs were either too opaque or required too much manual intervention. Shelvik built something that our buying team actually uses in their weekly review. The key was that they spent time understanding how we already worked before designing anything. That context made the difference.
January 2025 — Demand Planning Model
Case Studies
Three examples of how retailers have used Shelvik's services to address specific challenges in their e-commerce operations.
Challenge
Mid-size fashion retailer with 8,000 SKUs experiencing declining repeat purchase rates across their Singapore online store.
Their homepage and category pages served the same products to all visitors, regardless of browsing or purchase history.
Solution
Shelvik designed and implemented a personalisation engine drawing on 14 months of transaction and session data. The engine weighted recommendations by category affinity, price range, and recency of browse, with a fallback logic for new visitors. Integration documentation enabled the in-house developer team to handle deployment without ongoing Shelvik involvement.
Outcomes — 60 days post-launch
Click-through rate on recommended products
Repeat purchase rate (90-day window)
Total project duration
"What we found useful was that the integration documentation was written for our developer — not for a Shelvik engineer. We deployed it ourselves without any support calls."
— Digital Lead, Fashion Retailer
Challenge
Electronics retailer with seasonal peaks experiencing significant overstock in the post-peak period and stockouts during peak weeks.
Planning was conducted manually using last year's actuals, without incorporating promotional uplift or regional trend data.
Solution
Shelvik built a demand forecasting model using three years of sales data, incorporating promotional calendar events and a seasonal trend layer calibrated to Singapore's retail patterns. The model outputs a weekly demand view by category, delivered in a format the buying team could review directly in their existing planning spreadsheet.
Outcomes — Following Q4 cycle
Post-peak overstock volume
Stockout incidents during peak weeks
Total project duration
Challenge
Home and living e-commerce brand with a flat email engagement rate and a growing customer base it couldn't segment meaningfully.
All customers received the same newsletters. There was no differentiation between high-value repeat buyers and one-time purchasers.
Solution
Shelvik conducted a segmentation analysis on 11,000 customer records, identifying five distinct segments with different purchase rhythms, category preferences, and engagement patterns. The output was a written framework — naming and describing each segment — accompanied by concrete implications for email frequency, content type, and promotional sensitivity.
Outcomes — Next campaign cycle
Open rate for segment-targeted emails
Conversion rate from email traffic
Total project duration
Retail Clients
Avg. Client Rating
In Practice
Repeat Engagements
Ready to begin?
No pitch, no pressure. We'll ask about your situation and give you an honest view of what's possible.