Techbanq
Retail & E-commerce

Oasis Commerce Customer Intelligence

A unified customer intelligence layer for a multi-brand retail group, replacing four disconnected CRM and analytics systems with a single source of truth for marketing, merchandising, and store operations.

Client
Multi-brand retail group operating across 7 GCC countries
Category
Retail & E-commerce
Engagement
11-month engagement, phased rollout by brand
Services delivered
Data & AI systems, AI strategy & implementation, Digital transformation
The Challenge

The challenge

Group leadership needed a single customer intelligence capability that could serve four brands operating from separate data estates — but multiple previous attempts had stalled on the same structural problem. Four brands, three loyalty programmes, two CRM systems, and an e-commerce platform that did not connect to in-store point of sale meant that marketing ran campaigns blind, merchandising had no cross-brand customer view, and the store network had no real-time inventory visibility at cluster level. The technology investment had already been made; the capability was not materialising.

What we found

Discovery

Our diagnostic found that the systems were not the primary obstacle. The obstacle was that no agreement existed across the group on what a "customer" meant. The same person appeared with four IDs, three loyalty numbers, two email addresses, and no shared key. The first eight weeks of the engagement were spent designing an identity resolution model and getting the four brand CMOs to formally sign off on it — without that alignment, no technical solution would hold.

How we did it

Our approach

1

Identity governance design: designed the identity resolution model unifying customer records across loyalty programmes, e-commerce, in-store transactions, and digital touchpoints with explicit privacy and consent rules — and secured cross-brand leadership sign-off

2

Data foundation build: built a single source-of-truth customer event store capturing every transaction, interaction, and signal across all four brands and both channels

3

Analytical model programme: designed and built segmentation, propensity, and recommendation models on top of the unified data — each model owned by a named team, with retraining cadence and quality standards in the contract

4

Operational surfaces: delivered three production surfaces — a marketing activation console for the CMO offices, a merchandising dashboard at brand and cluster level, and a store-operations view reaching every shop manager in the network

What we built

The Outcome

The engagement delivered a unified customer intelligence layer giving the marketing, merchandising, and operations functions of every brand the same view of the customer — without forcing them onto a single tool. Each function consumes the data through the surface that fits its work, while the underlying identity model, event store, and model governance remain shared across the group.

Measured Impact

The results

+22%
Marketing-attributed revenue uplift in the first peak season post-launch
-38%
Reduction in duplicate customer records across the group
2.7x
Increase in cross-brand repeat purchase among unified customers
<48h
Time from new model to production for the marketing team
“For the first time, our four brands are operating from the same customer truth — without anyone losing autonomy over how they go to market. That was the political and technical problem we had been trying to solve for five years.”
G
Group Chief Customer Officer
Multi-Brand Retail Group
Technologies & PlatformsIdentity resolution engine with deterministic and probabilistic matchingCustomer event store on a managed lakehouseSegmentation, propensity, and recommendation model suite with weekly retrainingMarketing activation console, merchandising dashboard, and store-operations view
Next case studyHorizon Logistics Last-Mile IntelligenceA national parcel carrier replaces a 12-year-old rules-based routing engine with adaptive AI optimisation — collapsing per-stop cost and lifting on-time delivery in the same quarter.

Next step

Have a project like this?

We would love to hear what you are building. Let us explore how we can help.