Sukhmani Dhillon (Mani)

Data & Category Analytics · SQL · Tableau · Excel · Forecasting

I turn retail sales, inventory, and promotion data into decisions. Currently running category analytics for 110+ stores at WHSmith North America.

View my workContact me
Sukhmani Dhillon

94%

holiday sell-through

up from 78% on a forecast I rebuilt at store and SKU level

3.5x

vendor program income

year-over-year growth anchored by my ROI models

29%

multi-unit purchases

up from 18% in a BOGO test vs a matched baseline

96.4%

in-stock rate, top 100 SKUs

up from 91.2% after flagging chronic out-of-stocks

Results from my category analytics work at WHSmith North America.

Projects

Portfolio

Public-data projects, written the way I write for executives: findings first, methods after.

SQL

Live

Retail Vendor & Category Performance Analysis

I took a public e-commerce dataset (~100K orders, 9 relational tables, SQLite) and asked it the questions a merchandising team actually asks: where revenue concentrates, which sellers matter, and what operational levers move retention. Seven business questions, answered in SQL with CTEs and window functions.

  • Revenue has a long-tail problem: it takes 533 sellers (~18% of the base) to reach 80% of revenue
  • Late delivery carries a ~2-star review penalty: 2.27★ vs 4.29★ on-time, a 6.7x higher 1-2 star rate
  • Only 3.0% of customers ever reorder, but repeat buyers are worth ~2x (~$260 vs ~$138)
SQLSQLiteWindow FunctionsCTEs
View on GitHub

Tableau

Live

Retail Performance Dashboard

The same dataset, built into the kind of dashboard I build for executives at work: revenue and category KPIs, vendor concentration, regional performance, and the delivery-vs-review relationship in one interactive view, built on a star-schema data model with LOD expressions and table calculations.

  • $13M+ delivered revenue across ~96K orders visualized across 5 pages
  • Black Friday spike annotated: ~$988K month, +52% MoM
  • Star-schema model means every KPI has exactly one definition, the same discipline I use in production reporting
TableauLOD ExpressionsTable CalcsStar Schema
Live Dashboard

Excel

Live

M5 Demand Forecasting Engine

A full demand-forecasting build in Excel on Walmart's M5 dataset: 120 items × 4 California stores × 63 months (30,240-row fact table via Power Query). Five-model forecast ladder with a rolling-origin 4-fold backtest, seasonality indices, demand classification (ABC + Syntetos-Boylan), and error economics translating forecast misses into dollars.

  • Honest validation: expanding-window backtest across 4 folds, wMAPE/bias scorecard by model and demand class
  • Forecast Value Added waterfall: every model must beat the naive benchmark
  • Mirrors the seasonal forecasting I run at WHSmith (holiday sell-through 78% to 94%)
ExcelPower QueryLAMBDAForecastingBacktesting
Experience

Where the numbers come from

  1. Buyer / Category Analytics Manager

    WHSmith North America

    Jan 2025 to Present · Las Vegas, NV

    I run end-to-end analytics for the candy category across 110+ airport stores in the US and Canada: demand forecasting, promotion measurement, price testing, and vendor ROI modeling, built on SQL and a Tableau reporting layer I designed and own on Tableau Server. Recent highlights: holiday sell-through up from 78% to 94% on a forecast I rebuilt bottom-up at the store and SKU level, a 3.5x year-over-year increase in vendor program income anchored by my ROI models, and an executive dashboard that retired five manual workbooks and became the CCO's Monday-morning read.

  2. Associate Buyer / Category Analyst

    WHSmith North America

    Dec 2023 to Dec 2024 · Las Vegas, NV

    I built the department's first vendor ROI models and used them to pilot measured-lift funding on the specialty program, growing its income 13x in one year while program sales grew more than 70%. Along the way, my weekly SQL exception reporting surfaced the velocity decline that became the company's first private-label line, and I automated the reporting cycle itself, turning a six-hour manual Friday build into a 25-minute refresh.

  3. Assistant Buyer / Category Analyst

    Ross Stores

    Sep 2022 to Nov 2023 · New York, NY

    My first analytics seat: weekly demand forecasts across a 500-SKU assortment, where I built the team's first forecast-accuracy tracking and brought rolling MAPE from 34% to 26%, armed cost renegotiations with landed-cost and sell-through analysis, and proved a packaging-standards change cut damage rates by roughly 40% using an adopter versus non-adopter comparison.

B.S. Business Analytics & Marketing, Indiana University, Kelley School of Business, GPA 3.97 · M.S. Computer Science (Machine Learning & AI), WGU, expected Dec 2026

Skills

Toolkit

Query & Analysis

SQL (CTEs, window functions, multi-source joins), Advanced Excel (forecast models, OTB and retail math: sell-through, WOS, inventory turn, GM/AUR, Power Query, pivot models), MS Access

BI & Visualization

Tableau (executive dashboards, calculated fields, star-schema data models, KPI reporting), data visualization, Git/GitHub

Forecasting & Planning

Seasonal demand forecasting, MAPE/bias tracking, store-level allocation, consensus demand reviews with supply chain, replenishment parameters

Promotion & Pricing Analytics

Promo pre/post measurement with matched baselines, trade funding ROI by SKU, price-point testing, basket and transaction-log analysis

Retail Systems

Aptos Analytics, Merchandising Analytics (MA), Oracle RMS/RDW, Celigo

About

A category manager who runs on analysis

I'm a data analyst who grew up inside retail P&Ls. Today I run category analytics for a North American travel retailer: forecasting, promotion measurement, pricing tests, and the Tableau reporting layer executives actually use. I like problems where the data is messy, the stakes are commercial, and the answer has to survive a negotiation.

B.S. in Business Analytics from Indiana University's Kelley School of Business (3.97 GPA). Currently completing an M.S. in Computer Science with a Machine Learning & AI concentration at Western Governors University, expected December 2026.

Contact

Get in touch.

Email LinkedIn GitHub Tableau Public