Illustrative example. This scenario reflects the kind of project we deliver. It does not describe a named client, and details have been generalised.
Challenge
The retailer operates stores and an online shop in the UAE and Saudi Arabia, competing with leading marketplaces, specialist electronics chains and brand-owned stores. Its pricing team checked competitor prices manually, visiting websites for a list of priority products each week. The process was slow, covered only part of the range and could not keep pace with marketplace listings that changed several times a day.
Three problems stood out. Matching was inconsistent, so the team sometimes compared different storage capacities or model years. Promotional prices were recorded as regular prices, which distorted the picture during sale periods. And prices differed between Dubai and Riyadh, but the manual checks did not capture location.
Approach
We began with a workshop to understand the decisions the data would support: weekly range pricing, promotional response and a price-match policy for store staff. From this we agreed a tiered scope. Key value items, such as flagship smartphones, laptops and televisions, would be captured several times a day. The wider range would be captured daily.
Product matching was built on model numbers and manufacturer part numbers where available, with brand, capacity, colour and region variant as confirming attributes. Each match carried a confidence score, and low-confidence matches were reviewed by an analyst before they entered the feed. Bilingual listings were normalised so that Arabic and English product names for the same model were linked.
For each offer we captured the selling price, the reference price, promotion text, seller and fulfilment type, stock status and delivery promise, collected from set locations in the UAE and Saudi Arabia. Source terms were reviewed before collection and request rates agreed per site.
What we delivered
- A matched competitor price table covering the agreed competitors in both countries, refreshed daily and several times a day for key value items.
- Separate fields for regular, promotional and member prices, with promotion mechanics recorded as text.
- Seller-level detail on marketplace listings, distinguishing first-party offers from third-party sellers.
- Location-specific prices for the UAE and Saudi Arabia, in AED and SAR.
- An exceptions file listing only products where a competitor moved beyond an agreed threshold.
- Delivery to the retailer's data warehouse, with a scheduled spreadsheet for the category team.
Every delivery passed schema validation, completeness checks against the agreed product list, anomaly detection on price movements and encoding checks for Arabic fields, followed by a manual spot review of key value items.
Outcome
The pricing team replaced its weekly manual checks with a daily, matched view of the market. Category managers now start each review from the exceptions file rather than from a blank spreadsheet, which shortened the time between a competitor move and an internal decision. Separating promotional from regular prices removed a recurring source of disagreement in pricing meetings, and location-level data made the differences between the two countries visible for the first time.
The retailer has since extended the scope to additional categories and asked for promotional calendars to be tracked ahead of major regional sale events.