← Back to work

Client work

Promotion and loyalty planning system

Two years of transaction data, working for the weekly promotion.

An analytical system that turns the retailer's own basket and loyalty data into next week's promotion plan and into personalised vouchers instead of blanket discounts.

Result
  • 300M+ transaction rows analysed
  • 2 years of product-level history
Industry
Retail & FMCG
Users
The retailer's commercial team
Project flow
  1. Transaction and loyalty dataBaskets joined to the loyalty base
  2. Basket analysisBuying patterns and customer groups
  3. Commercial reviewThe team makes the call on next week's plan
  4. Promotion and vouchersWeekly plan, plus personalised loyalty offers
Rhythm
The analysis runs inside the existing weekly commercial cycle.

What needed to change

The data was never the problem. Reaching it in time for the weekly decision was.

A grocery retailer sits on an enormous amount of information: every basket and every product, scanned across the whole store network, plus a loyalty base large enough to reveal real patterns by store, region and shopper.

The hard part isn't collecting it. It's turning it into next week's decisions.

What we built

01

Documented data layer

Purchase history is joined with loyalty data and documented, so the same analysis can be repeated week after week.

02

Weekly promotion planning

Promotions are grounded in what customers actually buy: buying patterns, basket structure and customer groups.

03

Personalised loyalty offers

Vouchers are issued to the shoppers they suit rather than discounted across the board.

What changed

Weekly promotion planning and personalised loyalty offers built on real purchase behaviour instead of intuition.

300M+

transaction rows analysed

2 years

of product-level history

99.99%

product master accuracy