Hourly Grocery Pricing Across 12 UK Supermarkets: Powering a Next-Generation Price Comparison App

Introduction

How a London-based consumer app went from manual price checks to 4.8 million product updates a day — with zero downtime in the first 180 days

Actowiz Solutions  |  Case Study  |  Industry: Consumer Apps / Retail Tech

Client Snapshot

The client is a London-headquartered consumer app helping UK households save on their weekly grocery shop by comparing prices across major supermarkets in real time. Backed by a seed round in 2024 and a Series A in 2026, the company set out to be the most accurate, most current grocery price comparison platform in the UK — a market with notoriously fast-moving prices and aggressive private-label dynamics.

Client name and exact retailer breakdown are anonymized at the company’s request.

The Business Challenge

In its first year, the app relied on a combination of manual price checks, weekly flyer ingestion, and best-effort scraping built by two backend engineers. As user growth accelerated past 380,000 monthly active users, three problems became existential:

• Price freshness was inconsistent. Some categories refreshed weekly, others daily, others not at all. A growing share of user-reported errors traced back to stale data.
• Coverage was patchy. Only seven of the twelve major UK supermarkets were tracked. Users repeatedly asked why Lidl, Aldi, M&S Food, Iceland, and Ocado were missing.
• Engineering capacity was being burned on infrastructure maintenance instead of product features. The in-house scraping stack broke an average of 3.4 times a week.

"Every hour of bad data was an hour of churn risk. We needed a partner who treated price data as a first-class product, not as a side gig." — Head of Product

Solution Architecture

1. Postcode-anchored crawler fleet

Actowiz deployed a distributed crawler fleet anchored to representative postcodes across all UK regions, capturing store-level pricing where retailers expose it (notably Tesco and Sainsbury’s) and chain-level pricing where they do not. Crawlers ran on residential IPs with mobile-app traffic patterns to maintain stability against anti-bot systems.

2. Loyalty-pricing capture

Clubcard prices on Tesco and Nectar prices on Sainsbury’s were captured as a second price field alongside the everyday price. This single capability differentiated the client’s app from every competitor in the market.

3. Real-time promotion engine

Multi-buy mechanics ("3 for £5", "2 for £1.50"), BOGOs, and percentage-off offers were parsed into a structured promotion schema so the client app could surface the cheapest equivalent basket, not just the cheapest unit price.

4. Self-healing schema monitoring

Each retailer module ran continuous schema validation. Any deviation from the expected response shape triggered an engineering alert before user-visible errors could spread.

5. Two delivery channels

A low-latency REST API powered the live in-app comparison experience. A nightly Parquet dump replicated the full dataset into the client’s data warehouse for analytics, personalization, and CPI-style historical analysis.

Results in the First 180 Days

• 4.8 million product price points refreshed per day, up from 380,000 with the legacy stack — a 12.6x increase in data throughput.
• Retailer coverage grew from 7 to 12 of the UK’s major grocery chains, with Lidl, Aldi, M&S Food, Iceland, and Ocado added in the first 10 weeks.
• Zero unplanned downtime on the production data feed in the first 180 days. Schema drift was detected and patched proactively in 14 separate instances before any user-facing error.
• User-reported price accuracy complaints dropped 71 percent quarter-on-quarter.
• In-app session length grew 28 percent, attributed by the product team to the addition of loyalty-card pricing and the new "cheapest basket" recommendation.
• Engineering hours redirected: approximately 1,100 hours per quarter previously consumed by scraper maintenance moved to product feature work.
• The client closed its Series A on the strength of the data moat — investors cited Actowiz-powered freshness and coverage as a differentiator in due diligence.

"Actowiz turned price data from our biggest operational risk into our biggest product strength." — CTO

Why It Worked

• Treating data as a product, not a script. SLAs, schema versioning, and proactive monitoring were table stakes — not stretch goals.
• Loyalty pricing as a differentiator. Capturing what other comparison tools missed gave the client a marketing story and a measurable UX uplift.
• Engineering offload. Letting the client team stop maintaining scrapers was as valuable as the data itself — it converted into product velocity.
• Promotion intelligence, not just prices. Surfacing the cheapest basket required structured promotion data, not raw price lists.